EKV 23/09

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

ANNELI CARLQVIST

STOCKHOLM Division of Heat and Power 9-Nov-09 Kungliga Tekniska högskolan 100 44 STOCKHOLM

Division of Energy Technology Chair of Heat and Power Technology Professor: Torsten H. Fransson

Title Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects Author Anneli Carlqvist Report No EKV .... Project title Turbopower (T6373) Pages Drawings ..... Supervisor KTH: Prof. T. Fransson Date 09-11-09 References ..... Assoc. Prof. A. Martin Overall responsible at KTH: Prof. T. Fransson Approved at KTH by: Signature: Overall responsible at industry: - Markus Jöcker Industrial partners: - Siemens, Energimyndigheten, VAC Approved by industrial partners: - Signature: - Approved for Restricted, to distribution list below: Open: X Abstract The European Union targets a reduction of greenhouse gas emissions by 20% of renewable energy sources in the EU energy mix by 2020. New generations of technologies have to be developed through breakthroughs in research if the vision is to be met. The deregulation of the electric power market has opened the market to investors seeking profits and growth via the installation of additional capacity. This literature review report is a introductory step towards a more detailed and targeted work for a system concept evaluation, analysis and optimization of solar turbines in concentrating solar thermal power production. Concentrating Solar Power (CSP) has grown in recent years to be the largest bulk producer of solar electricity in the world and every square meter of a CSP field that produces 400 - 500 kWh of electricity per year, saves 0,45 ton of CO2 and contributes to a 0,1 ton reduction of fossil fuels use annually. The CSP is divided in two concepts and three main technologies: Linear– and point focus concept; Parabolic trough -, Power Tower- and dish- Stirling engine technologies. Whereas the Power Tower technology has the best gross efficiency, the parabolic trough has the advantage to be the most mature technology, being in grid connected power plant operation for at least two decades. Dish- Stirling engine systems have the highest net efficiency but suffer from the sensibility to insolation fluctuations and lack of storage. There are various thermal storage medias and storage concepts to help achieving the best thermal characteristics for charging and discharging to the CSP technology in question. The dispatch time ranges from 0,5 hour for maintenance and service up to 16-24 hours thermal energy dispatchability. The most recently commercialized thermal storage media is phase change material, i.e. salt mixtures with good thermal characteristics. Research for other kind of media suitable for CSP-technology is performed continuously. The operation modes of a CSP plant depends on the choice of electricity production; dispatch electricity during day- time hours or prolonging the electricity production beyond sunset. The Organic Rankine Cycle (ORC) is another CSP concept using organic thermal fluid instead of steam to the turbine, at a lower temperature and pressure range. The power output is of range 1–5 MW for electricity production. The ORC ability to use simpler components and the possibility to deployment in rural areas as well as on limited space in urban areas, make it more economical feasible. The organic fluid is for example n-pentane and toluene. The net efficiency for ORC- plants has been proven to achieve 30% under optimal conditions, comparing to the Rankine cycle practical efficiency of 36-40 %. The experiences from CSP plant operation have raised a demand for improvement of the steam turbine’s work performance. The steam turbine has until recently being designed to be in continuous work. The steam turbine operating in solar thermal plants should have the capability to start up directly and go from idle to rated condition in a matter of minutes. The CSP-plant start-ups per year is higher where the cycling thermal stresses can cause material fatigue leading to shorter life time. Efficient collector, storage design and optimal turbine operation should reduce the number of start-ups and shutdowns per year and the transient loads on the steam turbine. Distribution list

Prof. T. Fransson HPT 1 Assoc. Prof A.Martin HPT 1 Dr. P. Almqvist HPT 1 HPT archive HPT 1 Markus Jöcker Siemens 1

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

TABLE OF CONTENTS

1 Introduction ...... 9

2 Electric Production from csp ...... 9

2.1 Global Market Initiatives ...... 10

2.2 Key-Technology Factors ...... 11

2.3 Steam Turbine ...... 13

3 aim and Scope ...... 13

4 Methodology and report structure ...... 14

5 Solar Concentrating Plants ...... 15

5.1 Parabolic Trough Collectors ...... 16

5.1.1 Primary circuit -Thermal oil ...... 18

5.1.2 Primary circuit thermal oil – Molten salt thermal energy storage ...... 20

5.1.3 Direct Steam Generation ...... 21

5.2 Linear Fresnel Reflector - LFR ...... 24

5.2.1 Compact Linear Fresnel Reflector - CLFR ...... 25

5.3 Power Tower - Central Receiver Plant ...... 27

5.3.1 Power Tower with direct steam generation ...... 30

5.3.2 Power Tower with molten salt thermal media...... 31

5.3.3 Power Tower – Atmospheric air ...... 33

5.3.4 Heliostat field layout ...... 34

5.3.5 Pit Power Tower ...... 35

5.3.6 Multi Tower Solar Array ...... 36

5.4 Circular Parabolic Solar Concentrator – Dish-Stirling ...... 36

5.4.1 Dish-Stirling System Characteristics, and Comparison to Other Technologies ...... 37

5.4.2 Dish Concentrator ...... 38

1

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

5.4.3 Dish-Stirling receivers ...... 40

5.4.4 Receiver Losses ...... 42

6 Thermal Storage for CSP ...... 43

6.1 Sensible heat and latent heat TES ...... 44

6.1.1 Sensible Heat TES - Media ...... 45

6.1.2 Latent heat TES – PCM’s ...... 47

6.2 Chemical Energy Storage (CES) ...... 50

6.3 Mechanical Energy Storage (MES) ...... 52

6.4 Thermal Energy Storage concepts (TES) ...... 52

7 power block features of Csp plants ...... 57

7.1 Steam turbine ...... 58

7.2 Dynamic Operation of Steam Turbine in CSP Plants ...... 62

7.3 Organic Rankine Cycle Plant ...... 67

7.3.1 Cooling Water Requirements ...... 70

8 Conluding remarks ...... 71

9 Table of selected current and projected CSTEPP ...... 71

10 Bibliography ...... 74

APPENDIX 1. Concentrating Factor – C 82 2. CSP Power Plants 85 3. TES – Media 89

2

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

LIST OF FIGURES

Figure 1. Example of a CSTEPP with parabolic trough and 2-tank molten salt passive indirect TES (SolarMillenium)...... 15 Figure 2. a) A semi-parabolic collector. b) A compound parabolic reflector ...... 16 Figure 3. a) Parabolic trough, Solnova, Seville, Spain, (Abengoa Solar, 2008)...... 17 Figure 4. Solnova-1 Seville, Spain, Abengoa Solar (Abengoa Solar)...... 19 Figure 5. Construction site of AndaSol 1-3 April 2008 (SolarMillenium) ...... 21 Figure 6. Recirculation in a DSG parabolic trough field with superheating. (Reproduced from ...... 21 Figure 7. Water injection in the superheating stage, DSG. (Reproduced from: Zarza, et al., 2005) ...... 22 Figure 8. Saturated steam DSG parabolic trough plant (Zarza, et al., 2006)...... 23 Figure 9. Example of a Fresnel reflector...... 24 Figure 10. An example of a LFR 1st and 2nd reflectors and absorber design – Solarmundo Fresnel Collector (Bockenamp, et al., 2003)...... 25 Figure 11. Ausra CLFR at Lidell Power Station (EPRI, 2008)...... 26 Figure 12. CLFR Power plant, Ausra concept. (Ausra, 2007) ...... 26 Figure 13. PS20 and PS10 behind, Plataforma de solar, Spain. (Abengoa Solar, 2009) ...... 28 Figure 14. Heliostat. (BrightSource, 2009) ...... 28 Figure 15. Example of a DSG Power Tower with a saturated steam receiver. (Solúcar Solar S.A, 2006) ...... 30 Figure 16. Example of a molten salt Power Tower with a molten salt receiver...... 32 Figure 17. Example of Atmospheric air Power Tower – volumetric air cooler. (Pietz- Paal et al., 2005) ...... 33 Figure 18. eSolar 46 MW Power tower concept. (eSolar, 2008) ...... 34

Figure 19. An Animated picture of the Luz II – concept, a 33 MWe solar cluster. (www.news.cnet.com) ...... 35 Figure 20. Binham Canyon Mine, US and an animated picture of a theoretical PPT design (DiBella, et al., 2009)...... 36

3

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 21. Examples of dish-Stirling concentrators, a) Science Application International Corporation, b) Western Governors’ Association (owned by SES 2005) ...... 37 Figure 22. Australian National University dish-Stirling concentrator...... 38 Figure 23. Examples of dish-Stirling concentrators. a) Stirling Engine System, b) Schlaich Bergermann und Partner ...... 39 Figure 24. MDAC-receiver (Andraka, et al., 1994)...... 41 Figure 25. Cummin/Thermacore heat pipe absorber. Capillary Wick structure including arteries to distribute the liquid metal (Andraka, et al., 1994)...... 41 Figure 26. Example of a volumetric receiver. (Diver, Richard B., et al.) ...... 42 Figure 27. Thermal Energy Storage based on storage media...... 44 Figure 28. Various PCM material/molten salts, thermal conductivity – Melting point relation. (Hoshi, et al., 2004) ...... 48 Figure 29. Chemical TES, Ammonia as media in a dish-sterling system. (ANU, 2008) ...... 51 Figure 30. Thermal energy storage based on passive and active systems...... 53 Figure 31. PCM-cascade TES concept for parabolic trough CSP (reproduced from Knutsson, 2008)...... 55 Figure 32. Example of a Solar thermal steam power plant for electricity production with no TES and fuel back-up...... 57 Figure 33. Example of Gland Sealing steam system...... 60 Figure 34. Annual direct Insolation period 2000 – 2008. (source: Abengoa Solar) ... 62 Figure 35. Diurnal operation of a CSP plant – TES storage, prolonged electricity production. (source: FLAGSOL) ...... 63 Figure 36. Diurnal operation of CSP - plant. TES storage as intermediate stage. (source Solar Millenium) ...... 63 Figure 37. Diurnal operation DSG CSP-plant for month of January (Arza, et al., 2004)...... 64 Figure 38. Diurnal operation DSG CSP-plant for month of June (Arza, et al., 2004). 65 Figure 39. Steam turbine diurnal operation in CSP. (source: Siemens Industrial Turbomachinery AB) ...... 66 Figure 40. ORC Tier-Cascade/Recuperator/Reheat steam turbine cycle with high- and low ...... 69 Figure 41. Schematic Diagram of Saguaro Solar Trough Plant Operation...... 69

4

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Appendix Figure 1. The solar point of view, the dilution of solar irradiation flux and the aperture factor from a solar collector point of view, where it sees the solid angle of the Hemisphere. 82 Figure 2. Solar irradiation onto a collector’s aperture and reciever 84 Figure 3. CSP plant for Direct superheated steam generation 86

5

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

LIST OF TABLES

Table 1. A comparison between different CSP technologies, 50 MWe, 3 hrs storage, reference systems, annual basis...... 12 Table 2. A selection of thermal oil for CSP primary circuit heat transfer fluid (Moens, 2004)...... 19 Table 3. Data comparison of selected dish Stirling systems, see previous pictures. (Mancini, et al., 2003) ...... 40 Table 4. PCM cascade salts used in TES for parabolic trough CSP. (Knutsson, 2008)) ...... 56 Table 5. Water usage of different plants based on data from US Department of Energy, (NREL 2006)...... 70 Table 6 a) Selected CSTEPP projects, operational and planned plants...... 72

Appendix Table 1. Solnova-plant, parabolic trough CSP, Seville, Spain. 85 Table 2. Solnova-plant, parabolic trough CSP Power cycle data, Seville, Spain. 85 Table 3. AndaSol 1-3, parabolic trough CSP, plant data, Spain. 85 Table 4. AndaSol 1-3, parabolic trough CSP, Power cycle data, Spain. 85 Table 5. DISS/INDITEP Superheated steam DSG parabolic trough CSP data. 86 Table 6. DISS/INDITEP Superheated steam DSG parabolic trough CSP, power cycle data. 86 Table 7. DISS/INDITEP Saturated steam DSG parabolic trough solar plant 87 Table 8. DISS/INDITEP Saturated steam DSG parabolic trough solar plant. 87 Table 9. PS10-plant, Power Tower CSP, Seville, Spain. 87 Table 10. PS10-plant, Power Tower, Power cycle, Seville, Spain. 87 Table 11. Solar-Tres, SENER power Tower data, Spain. 88 Table 12. Solar-Tres, SENER power Tower data, Power cycle, Spain. 88 Table 13. Saguaro Organic Rankine Cycle, parabolic trough solar plant. 88

6

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Table 15. Thermal characteristics of TES liquid media. (Pilkington Solar Int., 2000) 89 Table 16. Thermal Characteristics of liquid salt mixtures. 89 Table 17 Thermal characteristics, TES – solid media. (Pilkington Solar Int., 2000) 90 Table 18. LHTS – Thermal characteristics of PCM salt mixtures. (Tamme R., 2003) 90

7

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

ACRONYMS

ANU Australian National University APS Arizona Public Service Co. CAES Compressed Air Energy Storage CCP Combined Cycle Plant CLFR Central Linear Fresnel Reflector CR Concentrating Ratio CRS Central Receiver System CSP Concentrating Solar Power DIR Direct Illumination Receiver DISS DIrect Solar Steam DLR Deutsches Zentrum fϋr Luft- und Raumfahrt DNI Direct Normal Irradiation The U.S. Department of Energy Solar Energy Technologies DOE Program DSG Direct Steam Generation GMI Global Market Initiative HP High pressure IAP International Action Programme Integration of DIrect steam generation Technology for Electricity INDITEP Production IP Intermediate Pressure LP Low Pressure MES Mechanical Energy Storage MCPAM MoleCular PhAse change Material MCPCM Microencapsulated Phase Change Material MTSA Multi Tower Solar Array ORC Organic Rankine Cycle PCM Phase Change Material PSA Plataforma Solar de Almeria R&D Research and Development ST Steam Turbine RTR Reversible Thermo-Chemical Reactions TES Thermal Energy Storage

8

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

1 INTRODUCTION

This literature review report is a introductory step towards a more detailed and targeted work for a system concept evaluation, analysis and optimization of solar turbines in concentrating solar thermal power production. Siemens Industrial Turbomachinery AB, Sweden in collaboration with KTH are the main stakeholders for the project “Steam Turbine Optimization for Solar Thermal Power Plant Operation”, financed by Siemens and the Swedish Energy Agency. It is a part of a long term industrial research effort in the field of thermal turbomachines and processes set with the overriding goal to contribute to the transition into a more sustainable energy society. Other interested parties within the field of sustainable development that could be interested in the overview are universities, environmental organizations and the solar thermal power industry. The overall aim of the report is to examine the system and component details of present solar concentrating thermal electricity power production plants.

2 ELECTRIC PRODUCTION FROM CSP

The European Union targets a reduction of greenhouse gas emissions by 20% and aims for 20% of renewable energy sources in the EU energy mix by 2020 (EC, 2007). It is stated in the European Strategic Energy Technology Plan (SET-PLAN) that in order to meet the targets, it is necessary to lower the cost of clean energy and put EU industry at the forefront of the rapidly growing low carbon technology sector . Furthermore acknowledgement is given to the fact that new generations of technologies have to be developed through breakthroughs in research if the vision to meet a reduction in the EU greenhouse gas emissions by 60-80% is to be met 2050.

The deregulation of the electric power market has opened the market investors seeking profits and growth via the installation of additional capacity. The competitive situation within this market contra the previous regulated environment has resulted in new set of key success factors as overall production cost, a power plant showing results competitive cost-wise and return of investments during the power plant’s lifetime. As a result the market for turbines has grown and also the competition

9

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

between established turbine suppliers and newcomers.

Electricity cannot be stored in a practical manner to meet the large scale and large fluctuation of the customers demand. Storing can be done in small quantities, (car batteries) or by indirect storage (f. ex. water dams and thermal energy storage) for large quantity electricity production. Therefore the electricity has to be produced in need-to-use basis or having the topography of large-scale storage. Also the demand for quick response time from the electrical generation plants when large fluctuations in demand occurs has to be considered in order to have a balance between demand and supply. CSP uses thermal energy storages to supply heat to the steam generation. The storage concepts avoid grid connected problems caused by other renewable sources of power production such as wind power or photovolatic.

2.1 Global Market Initiatives

Solar energy is the largest and most widely distributed renewable energy resource on our planet. Among the solar electric technologies, CSP is the lowest cost and the largest bulk producer of solar electricity in the world (Goswami et al., 2008). It is the ability to dispatch power when needed during peak demand periods that makes the CSP stand out from other renewable energy technologies and motivated development of solar thermal power plants. Implementing thermal energy storage systems that store excess thermal heat collected by the solar field during daytime enables production of electricity beyond day-time hours.

Potential technical feasibility faces the impediment of high initial deployment costs. This is also the case for CSP. The technology has not yet achieved mass production or optimization of its components and incentives are a key determinant of the rate at which CSP, as any new energy technology will be introduced in the early market phase. To face the problematic issues and reduce the barriers of the technology an international incentive started up by the Global Market Initiative. The GMI is an international public-private CSP partnership and a member of IAP, which during 2002 - 2003 facilitated and launched strategies for a rapid, large-scale market introduction of CSP (GMI, 2004). The initiative is based on agreements between government

10

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

organizations, CSP industries and third party organizations and aims for deploying 5,000 MW of CSP power by 2015 (Sargent & Lundy, 2003).

2.2 Key-Technology Factors

The preliminary requirements for the site of a CSP plant are sufficient direct normal insolation, water availability and electric transmission capability. The topography, especially in the case of parabolic trough technology, should not exceed 3 % inclination as it would imply higher cost for structure and anchoring to level the plant . The cooling technology will determine the amount of water usage – where dry cooling uses far less water than wet cooling and is suitable for more rigid areas to the cost of lower plant efficiency. Every square meter of CSP field that produces 400 - 500 kWh of electricity per year, saves 0,45 ton of CO2 and contributes to a 0,1 ton reduction of fossil fuels use annually (Goswami, et al., 2008). The energy payback time of concentrating solar power systems is less than one year. In addition, most solar field materials and structures can be recycled and used again for further plants.

Even though solar radiation is a source of high temperature and energy at origin, the insolation available for terrestrial use is much lower due to sun–earth geometrical constraints as well as the meteorological nature of earth climate that lead to a dilution of solar energy flux. Worldwide annual normal incident radiation for CSP varies from 2 1600 to 2800 kWh/m depending on the available radiation at a particular site. This rate assumes 2000–3500 annual full-load operating hours with the CSP (Goswami et al., 2008). Fluctuation in insolation like intermittent cloudiness results in rapid transients but can be mitigated by using an oversized mirror field and use the excess energy to load an energy storage system.

To overcome the constraints of decreased irradiance it is essential to maximize the collected solar thermal energy and minimize the heat losses within the CSP system when there is enough insolation during a satisfied time interval. The main requisite for solar thermal power plants is to have effective optical concentration devices and a thermal fluid transport system that increase the temperatures and the system efficiency. Therefore the solar collector field should have high-reflectivity mirrors, concentrating the incoming solar radiation onto a solar receiver with a small aperture

11

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

area. The solar receiver should have high-absorptance and transmittance and low- reflectance properties as well as negligible convection and conduction losses. To transfer maximum thermal energy to the thermal fluid transport system, CSP is divided in two concepts and three main technologies: Linear – and point focus concept; Parabolic trough/ Linear Fresnel lenses -, Power Tower with heliostats - and Dish Stirling engine systems.

Table 1. A comparison between different CSP technologies, 50 MWe, 3 hrs storage, reference systems, annual basis (Ortega, 2008).

Technology Parabolic Power Tower Power Tower & Trough Saturated molten salts Heat transfer Fluid, primary circuit Thermal oil steam Superheated steam Mean gross efficiency [%] Solar thermal to electric production 15.4 14.2 18.1 Direct radiation Mean net efficiency [%] Solar thermal to electric production 14 13.6 14 Direct radiation Specific power generation 308 258 375 [kWh/m2year] Capacity factor [%] 23-50 24 Up to 75

CSP plants utilize a number of high-temperature working fluids in the primary circuit, such as thermal oils, molten salts, or steam. Table 1 shows a general comparison of CSP technologies and the primary circuit heat transfer fluid. It is based on reference systems, 50 MWe nominal power and 3 hours thermal storage, annual values (Ortega, 2008). The comparison gives the Power Tower concept with molten salts the best gross efficiency1. Even so, parabolic trough has the most mature technology, being in grid connected power plant operation for at least two decades, the strengths and weakness well known. Viewing the mean net efficiency2 there is not a large difference between the three concepts but the PT + molten salt system shows the highest values of specific power generation as well as the capacity factor3 which is favorable when developing future CSP.

1 Gross efficiency: without auxiliary electrical consumption integrated. 2 Mean net efficiency: Average efficiency including the auxiliary electric consumption. 3 Capacity Factor: Ratio of the energy produced during an interval of time to the energy the plant should have produced at full capacity. 12

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

2.3 Steam Turbine

Most CSP today uses a turn-key power block where the steam turbine manufacturer guarantees the operation’s conformance with the solar thermal plant operation. The CSP plant concept supports the steam turbine and the thermal storage by transferring the solar thermal energy to a thermal fluid that is pumped through a heat exchanger that transfers the thermal energy to water. The water is then evaporated and the steam produced drives a steam turbine generator delivering electricity to the grid.

The experiences from the operation of CSP plants and the production of electricity have raised a demand for improvement of the steam turbines’ operation performance. Until recently steam turbines have been designed to be in continuous operation – i.e. as base load turbine; using biomass, coal, oil or natural gas as energy resource. In the CSP-system concept, the demand on the turbine has increased dramatically; the number of starts per year is several times higher and causes heavy transient loads, risk of thermal stresses can cause material fatigue and life time estimation decreases. The steam turbines operated in solar thermal plants should have the capability to start up directly and to go from idle to rated condition in a matter of minutes. The start-up time Efficient design of the collector system and the thermal storage should reduce the number of startups/shutdowns per year and reducing the transient loads on the steam turbine.

3 AIM AND SCOPE

The objective of this survey is to provide an overview of the current state-of- the art of various CSP plant concepts and to present a brief introduction to the steam plant process and the steam turbine operation.

Limitations: The report does not consider specific technical details of the main components of the CSP data provided from industry are limited to a select number of established companies that are known to the market. Neither economical validation of CSP nor a discussion of the specific national political rules and regulations to implement the

13

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

technologies was carried out, except for occasional comments when fit into the discussion. The CSP plants considered are of size equal or larger than 1 MWe. The report does not imply to have a full overview of the CSP market – as it is a commercial field in rapid development many ideas, products and research are most certainly still to be compiled.

4 METHODOLOGY AND REPORT STRUCTURE

Information on CSP trends and developments has been obtained via research reports, articles, and Internet emanating from universities, government agencies and companies. Interviews with experts working in the field have given insight and vital information.

Chapter 5 gives an introduction to the different CSP concepts in use today and possible future deployment.

Chapter 6 contains a brief presentation on the water consumption demand of a CSP plant and shows a comparison between different cooling-systems and their water usage.

Chapter 7 continues with the system overview by giving examples of thermal energy storage and media, depending on the CSP technology.

Chapter 8 discusses the steam turbine and steam power plant in brief, aiming for providing a first overall knowledge of the complexity, and also includes the operation modes and the steam turbine behavior during stop/start phases related to the insolation fluctuations.

Chapter 9 gives concluding remarks the overall knowledge of the report.

Chapter 10 close the review by showing a list of selected CSP plants in operation and future start-dates.

14

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

5 SOLAR CONCENTRATING PLANTS

The solar field in a CSP plant concentrates the solar flux incident on the reflectors.

The geometrical concentrator factor, Cg, of a concentrating solar collector is based on the fraction of solid angle of the hemisphere that the reflector sees and reflects to the absorber/receiver. The maximum geometrical concentrating factor for a linear CSP system is 212 and for a point focus CSP system it is about 46 300. (see Appendix 1)

A concentrating solar plant can be divided in three major parts: solar collector field; thermal energy storage and power block with steam generator; steam turbine and condenser. Heat exchangers are in operation between thermal fluid/thermal storage and thermal fluid/steam generator when other thermal fluids than water is used.

Figure 1. Example of a CSP with parabolic trough and 2-tank molten salt passive indirect TES (SolarMillenium, 2009).

15

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

5.1 Parabolic Trough Collectors

Parabolic trough based CSP is a relatively mature commercial technology that has generated electricity to the grid for over 20 years. A parabolic trough construction reflects the solar radiation onto receiver vacuum tubes at the parabola’s focal point. Viewing Figure 2a), the parameter deciding the solar collector’s concentrating ratio besides the material, is the geometry of the reflector: the parabola radius vector [r] is dependent on the focal length [f], the angle [φ] between the optical axis and the radius vector according to following equation:

Eq. 1

The chosen geometry and material of the solar collector decide its concentrating ratio

a) b)

Figure 2. a) A semi-parabolic collector. b) A compound parabolic reflector

Another geometrical trough concept is the compound parabolic reflector (CPC) (Figure 2 b), based on two geometrical different parabolas that are tilted to each other with an acceptance angle of Θa, reflecting the incident solar irradiation onto a receiver at the bottom of the trough. The acceptance angle is the maximum angle where the incident solar rays enter the trough and distributed across the receiver surface. The interval of incident acceptance angle to the optical axis is (±Θa/2). Solar rays having larger angels will be deflected. Enlarging the acceptance angle by

16

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

reducing the height [h] will decrease the geometrical concentrating ratio [Cg] but allow a higher fraction of incident irradiation reaching the receiver (Brogren, 2004). The compound parabolic collector has reached the highest concentrating ratio of the two parabolic designs and is currently used in the “reflective tower” concept as a secondary reflector/receiver (Cordeiro, 1998). The CPC allows a reduction in receiver surface area and thermal losses, increasing its efficiency. Hereafter the term parabolic trough is synonymous to semi-parabolic.

The reflective surface of a parabolic trough is usually of thin film of silver or aluminum on a rigid support and has a nominal concentration factor, Cg, of 80 (EPRI, 2008). The thickness of thin film aluminum for parabolic reflectors has been reduced by half in later years from eight mm to four mm, without losing the optical properties. The receiver is a stainless steel tube with a selective surface ceramic coating to increase solar heat absorption, and to reduce irradiative heat losses. Today ceramic- metal absorption coatings have increased the amount of heat captured by the tubes to the point that plants using them produce 30 percent more power than the first- generation solar thermal demonstration projects of the early 1990’s (Fairley, 2008). The thermal fluids for parabolic trough system are synthetic oil, molten salts and water/steam.

a) b)

Figure 3. a) Parabolic trough, Solnova, Seville, Spain, (Abengoa Solar, 2008).

b) Parabolic trough (Reflechtech Solar).

17

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Each collector tracks the sun by rotation about one axis (horizontal). Figure 3 shows the working procedure of the sun tracking parabolic trough. The troughs are mounted in a north/south – direction which allows the single-axis structure to rotor the troughs in an east/west – direction along the path of the sun.

The trough plants have been typically designed with at least 30 minutes of thermal storage, primarily as a buffer against rapid transients (cloud coverage) that would cause sharp drops and rises in power output. This kind of TES stabilizes power output and allows system operators to arrange for alternative generation sources (EPRI, 2008). Trough plants with several hours of thermal storage capacity lead to better reliability and thermal energy dispatchability over the diurnal electricity production.

Depending of the specific regulations in the countries where CSP plant concept is of interest, different systems has been chosen as well as the limitation of the electricity power output. Typical capacities lie in the 20 - 150 MWe span with the overall size limited by national regulations and capital costs. Fossil fuel (oil or natural gas) auxiliary heat sources are often employed. The size of thermal energy storage varies, from one-hour “safe”-storage in case of maintenance or solar irradiation fluctuation, to several hours of dispatchable thermal energy for diurnal electricity production.

5.1.1 Primary circuit -Thermal oil

The thermal oils used today have enhanced thermal characteristics to meet the demand of adequate temperature interval of the solar collector field. Disadvantages are the detoriation of the oil, reducing the usable time; toxic and flammable nature which demand safety precautions during operation and maintenance. Table 1Table 2 shows commercial thermal oils with thermal characteristics.

18

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Table 2. A selection of thermal oil for CSP primary circuit heat transfer fluid (Moens, 2004).

Application Commercial

Fluid T ɳ viscosity Properties Product Type of oil [gr. C] [cP] Mineral oils (‐10) ‐ 300 Flammable Caloria Parraffinic HC's Synthetic oils 13‐395 4,98 / 0,22 Flammable Therminol VP‐1 Aromatic HC's Silicone oils (‐40)‐400 Expensive, Flammable

The Spanish company Abengoa Solar’s CSP concept Solnova-1 has 90 loops with 4 trough modules per loop. The thermal fluid chosen is a synthetic oil that indirectly generates superheated pressurized steam to a power cycle steam turbine. As seen in the Figure 4, the solar field supplies the Rankine cycle with heat of 400 ˚C, using the natural gas boiler as a back-up. The steam production to the turbine is induced by a number of heat exchangers using the thermal energy from the CSP plant and the Rankine cycle in the most effective way. The temperature will be approximately 390 ˚C /100 bar at the HP steam turbine inlet (see Appendix 2, table 1 & 2).

Figure 4. Solnova-1 Seville, Spain, Abengoa Solar (Abengoa Solar).

It can be seen in Figure 4, that the Solnova plant uses a reheat steam turbine that includes a high pressure- and a low pressure turbine. The exhaust steam from the HP turbine is reheated with thermal energy from the CSP side and redirected to the inlet of the low pressure turbine. The expanded steam from the LP turbine outlet is transported through the condenser, where the steam is condensed to water and

19

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

reused in the closed-loop. The auxiliary boiler is in parallel to the CSP plant. With a Rankin cycle thermal efficiency of 36 %, the annual solar thermal to electric efficiency becomes 19 % for the Solnova 1 plant. The expected CO2-savings are 31 200 ton/year.

5.1.2 Primary circuit thermal oil – Molten salt thermal energy storage

Another parabolic trough concept uses thermal oils with molten salt TES (see Figure 1). An example of such a concept is the AndaSol solar trough plants. Each plant represents a fully dispatchable capacity of 50 MWe. Each plant is expected to generate approximately 180 GWh per year and consists of 624 collectors 150 meters long. The thermal fluid is synthetic oil with operating temperature close to 400°C. At the AndaSol the steam power cycle is combined with 7,5 hours of full-load molten salt thermal storage (28 500 ton/tank) (SolarMillenium, 2009 ). Solar energy collected during the day will be transferred to a molten salt solution at a temperature of approximately 385°C (EPRI, 2006). More details can be seen in Appendix 2, Table 3 and 4.

Andasol 1 has a natural gas boiler in parallel to the solar field and the thermal storage. As can be seen in Figure 1, two storage tanks are included in the plant layout for steam production. Thermal energy from the solar field is transferred to the storage by a heat exchanger between the molten salt and the synthetic oil. Additional heat losses have to be accounted for as more intermediate devices are put into the system; other internal energy demands include temperature maintenance of the thermal energy storage system and electricity consumption of auxiliary equipment.

20

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 5. Construction site of AndaSol 1-3 April 2008 (SolarMillenium)

5.1.3 Direct Steam Generation

Knowledge and learned lessons regarding direct steam generation in parabolic trough system was achieved in the research project DISS and its follow-up INDITEP within the European community.

Figure 6. Recirculation in a DSG parabolic trough field with superheating. (Reproduced from

Zarza E., et al., 2006)

In a superheated DSG plant (Figure 6), each parabolic trough row is divided into three stages; water preheating; evaporation and superheating stage. The number of troughs in each row depends on the amount of thermal energy required to convert the water entering the solar field. The evaporation section and the inlet of the steam superheating section are connected by a compact water/steam separator, which in

21

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

turn is connected to a larger shared vessel (see Appendix 2, figure 3), from which the water is recirculated to the solar field inlet by the recirculation pump. To avoid cavitation a small fraction of cooling feedwater is mixed with the recirculation water at the suction side of the recirculation pump (see Figure 7), as the water pressure and temperature at the outlet of the recirculation vessel are close to saturation. The DSG concept demands operation stability of the solar field under uneven distribution of solar irrradiation and solar irradiation transients. Recirculating water flow also helps to overcome sudden reduction in solar irradiation transients by supporting an even temperature to the steam turbine. The vessel separates the water before the steam enters the steam turbine (Zarza, et al., 2005) (see Appendix 2, table 5 to 8, and figure 3).

Figure 7. Water injection in the superheating stage, DSG. (Reproduced from: Zarza, et al., 2005)

Stratification can occur when the water volume is differentiated by temperature zones. The feedwater flow per row for the solar field design point is a compromise between avoiding liquid-water stratification and pressure drop in the row. The mass flow of water in each evaporation row should be kept above a certain threshold depending on the pressure. The seasonal effect of winter/summer solar irradiation results in oversized parabolic trough field for summer operation. The flow has to be reduced at winter to collect enough thermal energy, but still remain above minimum level. To deliver the required steam conditions to the Rankine cycle the flow cannot be too high as the pressure drop will increase in the receiver tubes. Defocusing parabolic troughs sections during summer peak hours when the flow is too high is a solution but a waste of valuable solar thermal energy. The power block selected must be robust, and operable under flexible conditions in order to assure durability and reliability. The steam turbine has a maximum permissible overload, and once this

22

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

maximum is reached, some solar collectors must be defocused in order to not to exceed this value, which would require solar energy to be dumped in a heat sink that would require additional system components.

The saturated steam option for the DSG has a more simplified solar field layout than for the superheated concept as seen in Figure 8.

Figure 8. Saturated steam DSG parabolic trough plant (Zarza, et al., 2006).

The saturated steam option operates also in recirculation mode, having vessels to separate the water and the dry saturated steam before the HP steam turbine inlet at a temperature of 260-300 ˚C (Zarza et al., 2006). Before the HP outlet steam is allowed into the LP turbine there is another separation of water from the steam following a reheat of the saturated steam (not seen in fig.8), to reduce the moisture content. The saturated steam solar field option requires fewer components than the superheated solar field; both water/steam separators between the evaporating and superheating sections and water injectors controlling the steam temperature are eliminated. The number of parabolic trough collectors required for the saturated steam solar field is higher than for the superheated steam option. The larger solar field compensates for the lower design efficiency of the saturated steam option.

The highest fluid temperature in the collector field is the saturation temperature. When operating at DNI, the lower limit of solar collector field power input is 250 W/m2 due to the thermal losses of the collector field and the system’s parasitic losses (DISS/INDITEP projects). The solar field efficiency for the saturated steam option is 23

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

always higher than that of the superheated steam cause to the result of the lower fluid temperature in the solar field and therefore lower thermal losses. On the other hand the net efficiency of the power block is higher for the superheated steam process due to higher live steam parameters. (see Appendix 2, table 7 and 8).

The saturated steam option has the advantage of employing a water/steam separator at the interface between the solar field and steam turbine, acting as thermal energy storage if the turbine is operated with sliding pressure. The amount of saturated water and steam inside this vessel can feed the turbine with saturated steam for a few minutes in the INDITEP4 design when insolation fluctuations occur.

5.2 Linear Fresnel Reflector - LFR

A linear Fresnel solar collector field has the same design lay-out as a parabolic trough’s with aligned modules where the tubular receiver is mounted in the focal point of the Fresnel lens. The difference is in the reflector design - a Fresnel reflector is faceted on one side (see Figure 9), which allows a flatter plate surface design, giving a longer focal length and reducing the mass of the reflector material compared to a parabolic trough. The Fresnel reflector has a lower concentration factor due to the manufacturing process that produces facets that are less precise than the parabolic design. The nominal concentration factor, Cg, is less than 80 (EPRI, 2008).

Figure 9. Example of a Fresnel reflector.

4 INDITEP; Integration of DIrect steam generation Technology for Electricity Production. 24

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

The flat mirrors are carefully tilted and turned on their axes to reflect irradiation to the receiver, analogous to central receiver technology. The Solarmundo LFR is an example of a Fresnel concept with a primary and a secondary reflector; the first reflector is the mirror reflecting the insolation onto the absorber and the second reflector behind the absorber. The second reflector both enlarges the target for the Fresnel mirrors and insulate the absorber tube. The back of the second reflector is covered by a opaque insulation and the front as a glass plate that minimizes the convection losses (Bockamp, et al., 2003)

Figure 10. An example of a LFR 1st and 2nd reflectors and absorber design – Solarmundo Fresnel Collector (Bockenamp, et al., 2003).

5.2.1 Compact Linear Fresnel Reflector - CLFR

Compact linear Fresnel solar collectors have the option of directing reflected solar radiation to at least two absorbers in linear systems. This solution prevents blockage of adjacent reflectors and also the possibility to place the collectors close to the ground, minimizing wind load dependence and structure material usage. CLFR concentrates the solar radiation up to 30 times. The tower height is about 15 m and typical absorber lines 600 m (see Figure 11).

25

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 11. Ausra CLFR at Lidell Power Station (EPRI, 2008).

These reflectors have standard flat glass surfaces and the heat transfer fluid used is water (DSG) and a tracking system is used to allow optimum operation. An example of CLFR technology on the commercial market comes from the company AUSRA. Their technology is designed for saturated steam production up to about 285 °C at 70 bar (Ausra, 2007) (see Figure 12). Low-temperature operation, while less efficient, avoids many of the problems of high-temperature operation such as thermal losses and the need for more durable materials and components.

Figure 12. CLFR Power plant, Ausra concept. (Ausra, 2007)

The turbine’s outlet steam is cooled down to water and returned to the solar collector receivers. At Lidell Power Station, Australia a deep cavern is used to store hot water under pressure, where the pressure is contained by the rock and the overburden

26

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

weight (Mills, et al., 2004). The steam is flashed directly from the cavern into the steam turbine without any intermediate heat exchangers. A makeup water tank on the surface provides the system with additional water. Future use in stand-alone solar plants could be using low temperature turbines (EPRI, 2006).

The AUSRA CLFR solar collector concept has some advantages over the parabolic trough technology and dish-engine design such as:

* Ease of access for cleaning and maintenance demands are low. * Choice between steam supply, either direct to the power plant steam boiler vessel, or indirect via a heat exchanger. * Passive direct boiling heat transfer is used to avoid parasitic pumping losses and the use of flow controllers for the high temperature fluid collector loop. * The single-ended evacuated tubes is made of advanced all-glass evacuated tubes with low radiative losses. * The heat transfer loop is separated from the reflector field and is fixed in space, thus avoiding use of flexible high-pressure lines or high-pressure rotating joints as in parabolic troughs.

5.3 Power Tower - Central Receiver Plant

Central receiver plants, also called power tower, employ a collector field array of several thousand sun-tracking heliostats. Each heliostat has a solar tracker that redirects and concentrate sunlight onto the tower mounted single receiver, which transfers the solar energy to a thermal fluid. The most common reflective surface today is coated glass mirrors, where the nominal solar concentration ration of a Heliostat-field is between 800 to 1500 (EPRI, 2008). The receiver outlet temperature can reach up to 700°C depending on the receiver type (Pietz-Paal, et al., 2005). The reason for the height of the tower is to avoid shading or blocking of the solar radiation from adjacent heliostats. Field layout can vary depending on the reflector size, receiver and the geographical site. In the northern hemisphere the heliostats are placed in a half circular form around the tower on the north side to optimize the operation and minimize heat losses (see Figure 13).

27

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 13. PS20 and PS10 behind, Plataforma de solar, Spain. (Abengoa Solar, 2009)

A plant sizes less than 10 MWe, the heliostats are normally located to the North of the tower, whereas a plant size larger than 10 MWe, the heliostats typically surrounds the tower (EPRI, 2008).

Figure 14. Heliostat. (BrightSource, 2009)

28

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Heliostats have limitations in operations such as (Abengoa Solar, 2009a):

* The effective reflecting area of the heliostat depends on the cosine effect, meaning the half-angle between the incident solar ray on the mirror surface to the reflecting ray onto the receiver. Hence the geometrical area could be larger than the reflecting area. * Control of the heliostats concentrating operation onto the tower could, if not properly focused, cause structural damage via unintended hot spots. * Clean reflective surfaces are important as the efficiency is reduced, by build- up of dust or other contaminants if not properly cleaned. * In case of wind speed above 10 m/s the heliostats are set in stowed position to avoid structural damage.

The power tower concept comprises a number of central reciever technologies such as molten salt - , pressurized atmospheric air- , and saturated steam receiver. With each of the receiver types a thermal storage system can be incorporated, for example two storage tank for molten salt and steel drums for saturated water.

There are different types of receivers depending on the solar flux density. At high flux density the solar irradiation is focused directly on an arrangement of vertical tubes welded together to form bundles. The tubes are heated and by conduction and convection transfer the thermal energy to the thermal fluid system. Another receiver for the high flux option is the volumetric receiver. It consist of porous material, wire grid or foam, that absorb the solar irradiation within the material, increasing the effective area without increasing the heat losses, permitting the same power output to a reduced receiver area. At small flux density cavity receiver, containing non- welded vertical tubes can be used. The larger surface area needed is protected from convective losses by having a small cavity volume, reducing air movement. The Heliostat field area is however limited due to the small cavity opening, allowing only a limited acceptance area. The layout of the field is therefore depending on the height of the tower and the orientation of the cavity

29

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

5.3.1 Power Tower with direct steam generation

Examples of saturated steam Power Towers are PS10 and PS20 in Spain. PS10 is an 11 MW DSG Power Tower plant that started operating in 2007 at Solúcar Solar Park, at Sancùlar la Mayor in Seville (see Figure 15). See further information in Appendix 2, table 9 & 10. The steel drum has a pressure above the system steam pressure to ensure boiling of the water into steam when opening the valves. The higher pressure the thicker the steel drums have to be, therefore to optimize the drum material is important.

Figure 15. Example of a DSG Power Tower with a saturated steam receiver. (Solúcar Solar S.A, 2006)

The PS10 plant presented by Solúcar Solar S.A (2006) has 624 Heliostats, each heliostat of 121 m2 is tracking the sun in two axis. The heliostats are curve shaped arranged in 35 circular rows around the tower (see Figure 13). The receiver has a slant range where the focal point is at a distance equal to the slant range. The solar receiver is of cavity type, formed by four vertical panels of 5,4 m width x 12 m height. Each panel has the heat exchange surface of about 260m2. These panels are arranged into a semi-cylinder of 7 m of radius. The receiver is basically a forced circulation radiant boiler with low ratio of steam at the oulet, in order to ensure wet inner walls in the tubes. Feedwater around 50 ˚C vaporizes at 250 ˚C and 40 bar.

During operation at full load, absorber panels will receive about 55,0 MWT of

30

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

concentrated solar radiation with peaks of 650 kW/m2. Thermal storage comprises four pressurized steel tanks with a thermal capacity of 20 MWh, equivalent to an effective operational capacity of 50 minutes at 50% turbine workload, and also provides controlled temperature conditions during the steam turbine start/stop-cycles. Natural gas back-up is used to power production of 12-15 % of PS10’s capacity.

PS20 has the same design as PS10 but has the twice the capacity. The number of heliostats are 1 255 and the height of the tower is 165 m.

5.3.2 Power Tower with molten salt thermal media

According to the report from Pietz-Paal (2005), the molten salt strategy demands a high heat flux receiver to provide an outlet temperature of 560ºC to the Rankine cycle and the molten salt storage. With liquid salt as thermal fluid, for example nitrate salt, the hot storage tank has an operating temperature at 565°C and the cold storage at 290°C. The receiver thermal efficiency is said to be about 88 % at a return temperature of 290ºC. Steam is produced at nominal conditions of 125 bar and 540°C and reheat steam at a temperature of 540°C. The nitrate salt storage might have today a capacity between 3 and 16 hours of full load turbine operation. Additional heat needed to maintain a minimum salt temperature is provided by electric immersion heaters and resistance heat tracing, which are incorporated in the design of the storage tanks (EPRI, 2008).

31

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 16. Example of a molten salt Power Tower with a molten salt receiver.

Solar –Tres project is a 17 MW central receiver molten salt plant. A total of 2493 glass heliostats with an area of 120 m2 each will occupy 1,42 km2 of land area when the construction is finished. The molten salt thermal storage provides 16 hrs of 647 MWh of storage equal with 6250 ton high-capacity liquid nitrate-salt. The 43-MW steam generator system has a forced-recirculation steam drum. This design places components in the receiver tower structure at a height above the salt storage tanks, allowing the molten-salt system to drain back into the tanks. The power plant has a higher pressure reheat turbine, with high steam pressure and temperature conditions for relatively low size compared to conventional power plants (Ortega, et al., 2009.) The larger heliostat field and thermal storage is expected to enable electrical power generation 24 hrs a day during summer and have an annual capacity of 65 %- 71 % including the natural gas backup (EPRI, 2006). See Appendix 2, Table 11and 12 for further information.

For molten salt technology the receiver should be able to withstand rapidly change temperatures without being damaged, for example of transients in radiation where the temperature can change from 290 - 570 ˚C in less than one minute (Tyner, et al., 1996). Flow obstruction is an issue when using molten salt as a heat transfer fluid. The receiver can be subjected to constrained melting when parts of the panels is heated while other parts remain cooled, causing salt freezing. Another challenge can

32

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

be risk of tube rupture caused by lack of cooling due to pressure failure where the tube material is weakened by high temperature corrosion. The wind has a certain influence on the power tower both on the structure and during the pre-heating, increasing the convective losses. A forced recirculation evaporator that moves molten salt through the shell side of all heat exchangers reduces risk of nitrate salt freezing.

5.3.3 Power Tower – Atmospheric air

Figure 17. Example of Atmospheric air Power Tower – volumetric air cooler. (Pietz-Paal et al., 2005)

The atmospheric air receiver in Figure 17 works with a porous absorber for temperatures of 700 ˚C and above. The concept is able to produce steam of 480 – 540 ˚C at 35 – 140 bar to the Rankine cycle. (Pietz-Paal, 2005) The thermal storage (ceramic type with e.g. alumina pebble) reverses the air flow direction depending on charging or discharging mode. This concept has the advantage of short start-up time, proven to be within twenty minutes under stable operation conditions (Pietz-Paal, 2005). The disadvantage is the limitation of storage hours, therefore additional back- up burner and a TES hybrid design to increase the annual capacity factor would be needed. Another issue is the necessity to adjust the number of heliostats allocated to different aimpoints on the receiver to achieve the best thermal efficiency possible. Improperly adjusted aimpoints can cause beam spillage and non-optimal optical

33

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

geometry between the heliostats and receiver, resulting in elevated reflection and radiation losses as well as absorber overloading (Tyner, et al., 1996).

5.3.4 Heliostat field layout

While the heliostat field layout is typically semicircular (see Figure 13), other arrangements are possible for example eSolar’s solution consist of several units of heliostat field with numerous smaller dual tracking reflectors in a rectangular lay-out heliostat has an area of 1m2 (Floyd-associates, 2009). Cause to their small size, mounted low to the ground gives resistance to wind tensions. A 46 MWe power unit consists of 16 tower receivers with their individual two heliostat fields, covering a 64 ha land surface (eSolar, 2008). The units can be connected to scale up to customer needs generating from 46 MW solar electricity and above. The heat transfer fluid chosen is water/steam with DSG technology.

Figure 18. eSolar 46 MW Power tower concept. (eSolar, 2008)

Another layout is the Brightsource Power Tower system, developed within Luz II Tower Power concept from Luz parabolic trough solar plant (SEGS) in USA during the 1990’s. Brightsource’s heliostat field includes a numerous two-axis rotating glass reflectors (area of 14,4 m2) surrounding the receiver tower 360º. The solar receiver is a drum type radiant forced circulation air cooled steam boiler, and the thermal fluid is water/steam superheated steam at 550– 565 ºC at 160 bar (Brightsourceenergy, 2009). One “solar power cluster” is made up of a power tower with surrounding heliostats, generating 33 MWe which can be scaled up to for example 100 MWe, where three solar power clusters are in operation (Floyd-associates, 2009).

34

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 19. An Animated picture of the Luz II – concept, a 33 MWe solar cluster. (www.news.cnet.com)

5.3.5 Pit Power Tower

DiBella et al. (2009) presented a CSP concept utilizing abandoned open-pit mines, particularly those that are grid connected. Heliostats are placed on the terraced area and focus on the receiver across the mine. Several towers with high flux receivers could be included depending on the level of reflected insolation. The appropriate working fluids are molten salt or air; for the latter it is suggested that hot air could be used in a Brayton cycle, with possible coupling to geothermal heat storage. Figure 20 shows a schematic of this concept. Many decommissioned mining pits are filled with water that could be used for cooling and maintenance purpose. This concept is thought to be combined with an aero-electric facility that can benefit from the low- grade exhaust heat from the condenser. During the day, the super heated air deliver heat to the system, while the molten salt storages are charged in the mine below. At night, the molten salt and geothermal air from underground mining deliver heat to the system. The solar power tower with heliostats would fill a small portion of the main pit and heliostats in the parabolic shaped mine would not experience blocking or shading. Many open-pit mines in good solar locations with extended infrastructure throughout the world exist (see a) b) Figure 20).

35

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

a) b) Figure 20. Binham Canyon Mine, US and an animated picture of a theoretical PPT design (DiBella, et al., 2009).

5.3.6 Multi Tower Solar Array

The Multi-Tower Solar Array (MTSA) design uses two-axis tracking heliostats and multiple point focus receivers on many small towers. Some heliostats overlap another owing to the tower’s close proximity. In some regions of the heliostat field, adjacent heliostats will be directed to different aiming points on different towers. Hence, the MTSA increases the annual ground area efficiency and the annual reflector area efficiency. MTSA is suitable for temperatures of operation between 327 – 1027 ˚C, with temperatures between 727 – 911 ˚C being appropriate for small Stirling cycle turbines (APP, 15 April 2009). When using the concept of high concentrating MTSA, it could be possible to apply a beam splitting heliostats technology, i.e. splitting the radiation in two spectral parts with different photon energies: one for thermal use (low photon energy) and one for photovoltaic purpose (high photon energy) increasing the overall electricial output.

5.4 Circular Parabolic Solar Concentrator – Dish-Stirling

Another approach to achieving high concentration ratios is the dish concept, where a circular parabolic concentrator reflects direct normal insolation towards a focal point; a high-temperature receiver is integrated with a Stirling engine. The dish is mounted on a structure enabled with dual-axis solar tracking: azimuth elevation5 and polar tracking6. Dish-Stirling collectors are able to operate with a lower incident radiation

5 Azimuth-elevation tracking rotates the concentrator in planes parallel and perpendicular to the earth. 6 The concentrator rotates at a constant rate of 15 degrees per hour in a plane parallel to the rotation of the earth. 36

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

than other CSP technologies, 800–900 W/m2 (EPRI, 2008). Concentration ratios can reach high levels >13000 and is claimed to have the highest net efficiency for current CSP technologies, up to 27% (Fraser, 2008).

The dish-concentrator reflects the solar radiation to the receiver mounted at its focal point that converts the heat to electricity by a thermal fluid driving a Stirling engine. The engine’s working fluid should have high heat transfer capabilities at high velocities and high pressures (Fraser, 2008).

5.4.1 Dish-Stirling System Characteristics, and Comparison to Other Technologies

The Stirling engine converts thermal energy from the absorber to mechanical power by applying heat externally to a working fluid. The working fluid is contained within cylinders of the engine component in a closed loop. A regenerator inside the closed loop improves the efficiency by pre-cooling the working fluid as it moves from the expansion space to the compression space, and pre-heating the working fluid as it moves from the compression space into the expansion space. The compression space is cooled by a secondary loop where a refrigerant circulates through a common radiator with forced air cooling provided by a fan. Dish-Stirling Engines have the highest specific work output for any closed regenerative cycle but have a slower response to an increase or decrease in load. Further details concerning the mechanics of Stirling engines are beyond the scope of this report.

a) SAIC (source: Ecostar, 2004) b) WGA (source: Ecostar, 2004)

Figure 21. Examples of dish-Stirling concentrators, a) Science Application International Corporation, b) Western Governors’ Association (owned by SES 2005)

37

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

ANU Big Dish, 400 m2 (source: Ecostar, 2004)

Figure 22. Australian National University dish-Stirling concentrator.

According to Lovegrove, et al. (2006) the ANU (Figure 26) dish concentrator, with an 2 area 400m , is considerably larger than any other solar dishes produced elsewhere in the world. Even if the overall trend by other developers have been to increase size 2 over the years, 200m is the next biggest to have been built since 1994. The concentrator is composed of triangular mirror elements attached to the dish-frame, and delivers a peak concentration ratio of 1500. The receiver is a monotube boiler housed in a cross section cavity receiver that produces up to 100 g/s of steam that is superheated to typically 500°C at 4.5 MPa. Dish receivers of this kind could possibly provide steam at any temperature and pressure that commercially available steam turbines can work with.

A CSP dish’s optical efficiency is considerably higher than the trough or tower systems as the mirror continuously is controlled to receive the direct insolation, whereas the trough and tower suffer from a reduction in projected area due to a frequent low angle of incidence (cosine losses). Even so the dish systems share an almost identical range of analogous system components at the same level of development as tower systems.

5.4.2 Dish Concentrator

The parabolic concentrator must be designed large enough to transmit a significant fraction of reflected radiation from the concentrator on to the absorber. At the same time the size of the dish is constrained by structural limitations, including the weight

38

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

of the Stirling engine, receiver, and other components; moreover the dish must be rotatable. Commercial Stirling dish systems are commonly of the size 10 kWe and 25 kWe with an approximate diameter of the parabolic dish of 7.5 and 11 meters respectively (Figure 21and Figure 23) (Fraser, 2008). However, at the Australian National University a 400 m2 dish Stirling solar collector (Figure 22) has been recomissioned for commercial development and is now a subject for improvement of its components (Lovergrove K., 2006).

a) SES Suncatcher (Tessera Solar Inc.) b) SBP (Ecostar, 2004)

Figure 23. Examples of dish-Stirling concentrators. a) Stirling Engine System, b) Schlaich Bergermann und Partner The most durable mirror surfaces of silver/glass have reflectance ranges according to the presentation of Fraser (2008) between 91-95 %. A polymer reflective film that has optical properties of 94.5 % mirror reflectivity has been developed and the most innovative parabolic mirrors use stretched-membranes across a hoop or rim with a second membrane placed behind the first. A partial vacuum then pulls the first membrane into a parabolic shape.

The intercept factor of a dish Stirling engine system is the fraction of solar radiation reflected from the dish that enters the receiver aperture. It is influenced by the size of the aperture, the collector system, the collector rim angle7 and nonparallel sunlight. Increasing the intercept factor increase the fraction of the energy entering the aperture (Fraser, 2008).

7 Angle of the concentrator’s curvature. 39

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Table 3. Data comparison of selected dish Stirling systems, see previous pictures. (Mancini, et al., 2003)

Dish ‐ Concentrator SAIC SBP SES WGA (mod 2) Glass area [m2] 117 60 91 42,9 Focal length [m] 12 4,5 7,45 5,45 Rim angle [˚] 29 52 40 37 Concentrating ratio Peak 2500 12700 7500 >13000 Peak Capacity [kW] 23 8,5 25 8

The concentration ratio is the normalized value of the system’s peak concentration. The theoretical maximum is obtained at a rim angle of 45 degrees. A comparison of selected types of dish Stirling engines is shown in Table 3. Here the difference between the geometrical and optical parameters gives different concentrating ratios. An interesting comparison is between the SAIC and WGA models where it is evident that the size of the area and focal length has less influence on the concentrating ratio than the rim angle. WGA’s smaller size but larger rim angle (37˚) gives a concentrating ratio over 13 000 while the SAIC’s almost three times larger size dish has a concentrating ratio of 2500 at a rim angle of 29. The focal length shows the relation between the rim angle and the size of the dish. Larger dish with smaller rim angle gives a longer focal length.

5.4.3 Dish-Stirling receivers

The receiver consists of an aperture and an absorber where the aperture is located at the focal point of the parabolic concentrator to reduce radiation and convection losses. There are three kinds of absorbers (Fraser, 2008):

* DIR - Direct Illumination Receivers * Heat pipe absorbers * Volumetric absorbers

Current dish-Stirling absorbers are typically DIR and consist of a bank of tubes to directly heat the working fluid, using the solar radiation that is absorbed on the external surface of the tubes. DIR has the disadvantage of difficulties with balancing thermal input between the multiple dish-Stirling cylinders. The heater tubes will incur more thermal hot spots as compared with other absorber types.

40

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 24. MDAC-receiver (Andraka, et al., 1994).

The function of heat pipe absorbers is to vaporize a liquid metal such as sodium (675 ˚C) on the absorber surface and condense it on the dish-Stirling engine heater tubes to transfer the energy to the working fluid. Heat pipe absorbers yield more uniform temperature distribution on the heater tubes; thereby resulting in longer lifetime for the absorbers and engine heater heads in comparison to the DIR absorbers. (Andraka, et al., 1994; Fraser, 2008)

Figure 25. Cummin/Thermacore heat pipe absorber. Capillary Wick structure including arteries to distribute the liquid metal (Andraka, et al., 1994).

Volumetric receivers transmit solar energy through a fused silica quartz window, and absorb the energy onto porous ceramic foams, having the potential to operate at higher temperatures. One drawback is that the quartz window prevents approximately ten percent of the solar radiation from entering the receiver. Convection and radiation losses from the receiver are reduced, so the net effect of a reduction in radiation may be minimal.

41

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 26. Example of a volumetric receiver. (Diver, Richard B., et al.)

5.4.4 Receiver Losses

The majority of the overall thermal losses of a concentrator dish-system is generated from the receiver, responsible for up to 56-74%8, while the reflector losses due to the mirror reflectivity comprise between 24-37 % of the total losses. The conduction losses are minimal in the dish- Stirling receiver since they can be easily controlled by adding insulation without large losses in other components. The convective losses may represent up to 25 % of receiver losses during noon, and up to 40 % during the morning and evening. Convective losses are a function of cavity temperature and geometry, aperture orientation and diameter and wind velocity (forced convection). These losses are also dependent on the time of year and location since the angle between the sun and a horizontal surface changes with the season. The Radiation losses in the receiver may represent up to 60 % of receiver losses during the morning and evening, and about up to 75 % at noon (Fraser, 2008).

A 25kW dish-stirling collector needs only a landarea of about 510 m2, able to produce

55 - 60 MWeh/year. In both US and Spain plans for 25 kW dish-Stirling Engine clusters for grid deployment are under way. (Stirling Energy, 2008)

A comparison presented by Lovegrove (2006) between CSP technologies: parabolic trough, power tower and dish-Stirling, showed the annual overall insolation to electric efficiency, ɳte , to be highest for the dish-Stirling (14 %) and lowest for the parabolic trough system (11%), with the power tower concept close to the dish-Stirling’s efficiency. To be noted is the peak capacity for the dish-Stirling (1 MWe) system that

8 SES and WGA dish-Stirling systems. (Fraser, 2008) 42

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

is 30 times less than for the parabolic trough (30 MWe) and 14 times less for the power tower (14 MWe) for the stated efficiencies. The most significant differences between the technologies lay in the solar field optical efficiencies and receiver thermal efficiencies between the parabolic trough/power tower- and the dish-Stirling system, that are 30 % and 10 % higher respectively. More, the power cycle efficiency is highest for the power tower system (40%) and lowest for the dish-Stirling (27%) with the parabolic trough system efficiency of 35 %.

6 THERMAL STORAGE FOR CSP

Operational flexibility and economic feasibility of CSP plants depend a great deal upon thermal energy storage (TES) approaches. The TES storage media has to meet a number of essential constraints: high energy density; high thermal stability (exceeding 425 C); low or non-toxicity; non-flammability; compatibility with other materials (non-corrosiveness); and low cost. Mechanical and chemical stability of the container are equally important as compatibility between the heat transfer fluid, heat exchanger and storage media. The TES-system has to be able to operate within the solar fluctuation behavior and be well integrated into the power plant, having optimal operation parameters and strategy with minimum heat losses to the surroundings during multiple charging/discharging cycles. Research is continuously done to develop different storage systems as well as the heat transfer fluid that can be used in both the solar field and the storage system. The thermal conductivity of the media gives information about the media influence on the heat transfer design and the heat transfer surface requirements on the storage system. Designing the thermal plant aims for high volumetric heat capacity improving the TES by reducing the size, external piping and structural costs. The most common thermal storages in CSP commercial use are of pressurized steel drums for steam or water and molten salt storages.

43

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

TES systems can be classified according to the storage media system or the storage concept as seen in Figure 27 and Figure 30.

- Sensible Heat Compressed Air - Latent Heat Mechanical Energy Thermal Energy

CSP TES MEDIA SYSTEMS

Figure 27. Thermal Energy Storage based on storage media.

6.1 Sensible heat and latent heat TES

Sensible heat means changing the internal energy of a media without changing its phase. The material temperature increases and thermal energy is stored until discharging by decreasing the temperature. The important properties of a sensible heat storage material are the ability to conserve heat per unit mass9, and the material average density as well as operational temperature, vapor pressure, compatibility among materials, stability and surface area/volume ratio and heat loss coefficient.

Thermal energy can be stored by latent heat of phase change in some media. The latent heat energy absorption is a reversible isothermal process between two phases. When a phase change from solid to liquid state occurs, the material absorbs energy in the melting process at an almost constant temperature and releases energy in the reversible transition back to solid phase. The melting process is limited by thermal conductivity and natural convection, and the solidification process is mainly limited by thermal conductivity. Other phase change states are solid to solid crystalline phase transformation and liquid to vapor. Due to the isothermal process the latent heat storage has the advantage over sensible heat storage systems, where the temperature drops as the storage is discharged. (Hoshi. et al., 2004). Also beneficial

9 Specific heat capacity. 44

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

is that the vessel containing the latent energy media is not exposed to significant thermal stresses and can be reduced in size compared to the single-phase sensible heating systems. A solid-solid phase change, that some salts exhibit a major stress effect on the storage container at multiple cycling (Hoshi. et al., 2004). Sodium chloride as a solid media has a heat capacity per unit mass comparable to concrete (kW/kg) but 50% less heat capacity per unit volume (see Appendix 3, table 17).

6.1.1 Sensible Heat TES - Media

Gas media must be stored under pressure to maintain sufficient amount of heat at a certain volume and requires large storage volume. Multiple steel drums are used at CSP-plants where this concept is chosen mainly where DSG in the solar field can drive a steam turbine directly. Liquid media for sensible heat storages are thermal oils and various salt mixtures. Thermal oils are liquids at room temperature and thus a more convenient to handle, but some are environmentally questionable due to their toxic nature. Salt mixtures are solid, having a higher freezing point up to 200 ˚C (Knutsson A, 2008). A higher freezing point gives higher technical and economical limitations when it comes to auxiliary heating during off-operating hours.

Thermal oils are essentially petroleum based mineral oils for TES-systems up to 350ºC (See Appendix 2, table 15). Thermal oils becomes unstable above 300 – 350 ºC (Knutsson, 2008), where they change properties, so they are inappropriate for high temperature storage, i.e. Power Tower. Synthetic mineral oils with improved properties and higher temperature range are now commonly used in the CSP system today, for example parabolic trough systems (see Table 2).

Salt mixtures cover the temperature range from 120-1400 ºC; the thermal characteristics increasing with increasing temperature (see Appendix 3, table 16).

Today the nitrates (-NO3), nitrites (-NO2) and carbonate (–CO3) salts are under consideration for TES use (Knutsson, 2008). Mixtures of eutectic inorganic salts with wide temperature range like 60 % potassium nitrate and 40 % sodium nitrate have proven to be satisfactory as TES media with a freezing point of 221ºC.10 An eutectic mixture called HITEC made up of nitrate salts (NaNO3, 7% / KNO3, 53%) and

10 Also called ”Solar salt”. (Knutsson A, 2008) 45

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

potassium nitrite salt (NaNO2, 40%) have proven a lower freezing point of 142 ºC, due to the nitrite salt lowering the freezing point (Pilkington Solar Int., 2000). A third salt called HITEC XL combined of 48 % Ca(NO3)2, 7 % NaNO3 and 45 % KNO3 has shown a freezing point interval between 87 – 130 ºC (Kearney, et al., 2001). Thus the storage material is more resistant to unwanted freezing in the case of power plant shutdown or long time of no insolation. In general fused salts are highly corrosive and inhibitors are required to prevent corrosion. Other examples of salts useful as thermal storage media are: phosphates (-PO4), oxides (-O), sodium chloride, magnesium chloride and aluminum chloride. See Appendix 3, table 18 for more information

The carbonates have the best heat values of all salt types (see Figure 28; Appendix 3, table 15). Having the best volume specific heat capacity, the carbonates are able to store energy at one third the space nitrites would need for storage of the same energy amount, and about at half the space of nitrates. More, nitrate salt has higher specific heat capacity per unit volume than the nitrite salt and is able to store 50 % more thermal energy (Knutsson, 2008). Nitrates are highly corrosive at high temperatures. To reduce the corrosive effect, nitrite can be added (Knutsson, 2008).

Liquid metals like sodium and sodium-potassium have properties similar to water, good heat transfer characteristics and have high temperature ranges while maintaining the thermal storage capacity and stability. A liquid sodium metal storage tank would need less volume than for the thermal oil to store the same amount of thermal energy (see Appendix 3, table 15). This is due to the floating sodium which compared to synthetic mineral oil has adequately higher volume specific heat capacity. Liquid metals could be considered for high temperature TES applications, particularly where high heat transfer rates are needed. They have however high reactivity with container material, water and air as well as low specific heat properties11 (see Appendix 3, table 15). When properly contained and maintained the liquid metals can minimize the corrosive impact.

11 For high temperature sensible TES-systems, liquid metals has an average heat capacity. (Knutsson A., 2008) 46

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Examples of solid media usable for thermal storage is castable ceramic, concrete, sodium chloride, cast iron, cast steel, silica- and magnesia fire bricks (see Appendix 3, table 17). A castable ceramic has lower material strength compared to concrete but has better durability and resistance to cracks. 12 There is also a thermal expansion to regard when calculating the storage container. By comparison magnesia brick has the highest heat capacity yet low heat conductivity. Cast steel and cast iron has up to 40 times higher thermal conductivity than other tabulated solid materials except for solid sodium chloride and magnesia bricks (see Appendix 3, table 17). Most solid metals are not relevant for TES applications, consider low heat of fusion per unit weight or volume, unfavourable chemical properties and in some cases toxicity (Knutsson, 2008). Solid media heat storage could be an attractive option for parabolic trough power plants that uses synthetic oil as the heat transfer medium. Two storage systems of castable ceramic and high-temperature concrete have been tested at PSA in Spain (EPRI, 2006). The two concepts proved suitable, had integrated tubular heat exchanger between the thermal oil and the storage media. The high-temperature concrete would be the superior thermal storage cause of its lower cost, higher material strength, and easier handling.

6.1.2 Latent heat TES – PCM’s

Utilizing the latent heat involved in a phase change of a media is the definition of the phase change materials (PCM) technology. During liquid to gas transition, the latent heat media in gas phase has high energy per unit mass of the storage medium, thus less mass of storage material is necessary. The disadvantage with liquid to gas phase change media is the average heat capacity per volume is small and requires large volume of the storage container due to the gas-phase. In CSP plants where the heat transfer fluid in the solar field contains a two-phase media water/steam, multiple steel drums are used as thermal storage.

The most interesting PCM option for intermediate and high temperature storage is the melting-freezing phase change of salt compositions such as nitrates, hydroxides, salt ceramics, and chlorides. For a PCM to be a relevant choice for parabolic trough solar power plant storage, the phase change temperature of the

12 Thermal expansion Coefficient 11.8, [10-6/K], Castable Ceramics, DLR 47

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

material should be in the interval suitable the storage to provide the Rankine cycle with thermal energy.

Studies of PCMs as storage element in a parabolic trough system13 shows that some salt mixtures like NaOH/Na2CO3 have a superior heat capacity to single salts of similar melting point such as NaNO2 (see Appendix 3, table 18). Above this temperature and up to 427 ˚C, metals such as Sn and Zn can be used as PCM’s (Hoshi, et al., 2004).

For Fresnel systems according to Hoshi, et al. (2004) and Tamme, et al., (2003) using low pressure turbines at or below 327˚C and 25 bar, the salts NaNO2, NaNO3 and KNO3 are interesting PCM’s with saturated steam as heat transfer fluid. Figure 28 and table 18, Appendix 3, show PCM’s with phase change temperatures reaching 900 ºC. Mixtures of these salts as main ingredient give different suitable phase change temperatures for TES systems.

Figure 28. Various PCM material/molten salts, thermal conductivity – Melting point relation. (Hoshi, et al., 2004)

13 Thermal oil as heat transfer fluid. 48

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Viewing Figure 28, for MTSA-technology (see section 5.3.6), high melting point media with high thermal conductivity such as Na2CO3 should be of interest in the latent heat TES-systems. It should be noted that in general the melting point and the heat capacity tend to increase in the order of nitrates, chlorides, carbonates and fluorides.

Three generic configurations for PCM thermal storage systems have been identified in the report by Luz Int., et al., (1988): - Shell-and-tube heat exchangers. PCM contained in the tubes while the heat transfer fluid circulates on the shell side or vice versa - Packed bed configuration with encapsulated PCM spheres contained in a pressure vessel. - Direct contact configuration between the PCM and the heat transfer fluid

The storage systems utilizing latent heat PCM’s can be reduced in size compared to single phase sensible heating systems. This is due to the relative high latent heat of fusion between the liquid and the solid state. There could however be a limit to the number of freezing and melting cycles where degradation of the PCM material reduces the performance of the TES storage (Zalba et al., 2002). In the intermediate and high temperature ranges, there is also a risk of corrosive effect of the phase change material on the container. More, the low heat conductivity prevents the full use of PCM storage technology, but the heat transfer in the storage tanks can be improved choosing the PCM in such a way that its phase change temperature optimizes the thermal gradient with respect to the opposing heat transfer media. The maintenance of the storages has an elevated risk of high parasitic loads due to daily processes of filling and emptying the receiver and ducts (Solúcuar Solar S.A, 2006).

There are efforts to develop the sensible and latent TES medias to achieve a high temperature, low vapor pressure, low melting point heat transfer fluids that would eliminate the weaknesses of both the synthetic oils and molten salts, while simultaneously providing both short- and long-term thermal energy storage. Examples of other possible heat transfer fluids are: molecular alloys phase change

49

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

material (MCPAM) (Zalba, et al., 2002) that are thermo-adjustable14; presented by Moens (2004) are plasticizers, modified high temperature lubricants such as esters and modification of commercial fluid such as Therminol VP-1TM ; nanoPCMs15 (Glatzmaier, et al., 2008) and microencapsulated phase change material (MCPCM) (Yamagashi, et al., 1998). Introducing the MCPCM particles in the collector loop, the heat transfer fluid could serve for a short-term buffer storage as the effective specific heat of the heat transfer fluid increases. The particles might also be separated from the heat transfer fluid and stored in a vessel as a long term storage, until needed for discharging at which point they would be reintroduced into the heat transfer fluid stream. The potential risks are erosion of pumps, valves and other collector loop components, as well as difficulties of encapsulation and the complexities of the separation and storage phases. Research on ‘room temperature ionic liquids’ such as imidazolium salts has shown very high thermal stability (over 400˚C), almost no vapour pressure, low melting point and are non flammable although these salts are thermally unstable for typical CSP-technology temperatures (Moens, 2004). In Appendix 3 tables 15 through18 of the discussed TES media can be viewed.

6.2 Chemical Energy Storage (CES)

Chemical storage is storing thermal energy by utilization of the energy in reversible endothermic/exothermic chemical reactions. The solar thermal energy excites an endothermic chemical reaction that should ideally be completely reversible (Pilkington Solar Inc., 2000). Since it is important to control the thermal energy exchange, catalyst induced exothermic chemical reactions are favorable. This process is called reversible thermo-chemical reaction (RTR). The advantages are high storage energy density, long storage duration at near ambient temperature and heat pump capability. Unfortunately CES media show uncertainties in the thermodynamic properties and the reaction kinetics as well as being toxic and flammable. CES media such as electrochemical flow batteries capable of storage times for several minutes or hours have been demonstrated at utility scale but not yet in conjunction with solar energy systems (EPRI, 2006).

14 The material’s composition allow temperature phase change alterna tion. 15 Nanoscale “encapsulated” structures (50 – 500 atoms) capable of phase transitions. 50

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Ammonia as a heat transfer fluid and thermal heat storage, that has been chemical produced in well-developed, large scale systems has received attention as a potential storage medium for CSP. Chemical storage of ammonia has an operation temperature from 0-700ºC (Knutsson, 2008) and could thereby be applied to a CSP system and it could be possible to use ammonia as heat transfer fluid, reducing the components and the heat losses. Using the concept of ammonia in a closed loop, reactants in form of ammonia, nitrogen and hydrogen pass from energy storing to energy releasing reactors in an ammonia synthesis reactor.

Figure 29. Chemical TES, Ammonia as media in a dish-sterling system. (ANU, 2008)

Figure 29 shows a dish-concentrator with ammonia TES as proposed by ANU (2008). In the endothermic process ammonia is dissociated while absorbing heat, producing nitrogen (N2) and hydrogen (H2). When extracting the heat from the exothermic process the ammonia is resynthesized. The reactor contains catalysts to promote the reversible process. There are minimum of heat losses due to the fact that the solar energy is stored in chemical form at ambient temperature. The ammonia has a high energy storage density both by volume and mass. A temperature of 400 - 500 ˚C can be achieved, reducing the thermal losses from dish- receivers and material cost by cheaper optics and avoiding some of the high temperature materials limitations otherwise an issue. One storage concept is required as the ammonia is liquid in ambient temperature while the nitrogen and hydrogen remain in gas phase. See ANU ( 2008) for more information.

51

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

6.3 Mechanical Energy Storage (MES)

Compressed air energy storage (CAES) is a relatively mature large-scale energy storage technique. CAES uses electrically driven compressors to charge an underground reservoir during off-peak hours. When needed, air is discharged from the reservoir into an expansion turbine connected to a generator. Alabama electric cooperative built a 110-MW CAES facility in 1991 with EPRI participation and more recent projects have proposed taking advantage of salt caverns in Texas and limestone mines in Ohio. (EPRI, 2006)

6.4 Thermal Energy Storage concepts (TES)

Thermal energy storage concepts can be divided into active and passive storages (Gil, et al., 2008) (see Figure 30). Both active and passive storages can be subdivided in direct and indirect storages. Direct indicates the solar field heat transfer fluid is also used as the storage medium. In an indirect storage system the solar field heat transfer fluid is separated from the storage medium via a heat exchanger. Direct storage systems eliminate losses associated with the heat exchanger used in indirect systems. However, some solar field heat transfer fluids have pressurization requirements that would make direct storage systems too expensive, where indirect system design shows the flexibility by using different fluids in the solar field and storage system. Thermal energy storage approaches include the single tank thermocline (in which hot and cold heat transfer fluids are stored in the same stratified tank), the single tank trickle-charge thermocline (in which the tank contains a large volume fraction of low-cost sensible heat storage media and the heat transfer fluid is trickled over the media), the two-tank molten salt systems and the dual media systems (in which the heat transfer fluid is pumped through pipes imbedded in the storage media.

52

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

CSP TES CONCEPT

Active System Passive System

Direct Indirect Direct Indirect

-Direct Steam - Two tanks - Single tank Thermocline Dual media - Cascade - Single tank Trickle - Concrete Charge Thermocline - Ceramics

Figure 30. Thermal energy storage based on passive and active systems.

Two-tank thermal energy storage systems use two reservoirs, a hot tank and cold tank, to store thermal energy. To charge the direct system, high temperature fluid exiting the solar field is accumulated in the hot tank while low temperature fluid is pumped from the cold tank to the solar field. When discharging, the fluid in the hot tank is pumped to provide thermal energy to the power cycle and before returning to the cold tank.

Molten salt storage is a proven technology that has been used in the process industry. Andasol-1 molten salt 2-tank is a liquid TES (see Figure 1), classified as a passive indirect storage concept. The storage capacity is totally 1010 MWh thermal energy (Nava, et al., 2006), constructed for 7,5 hrs dispatchability. Two separate tanks are connected to the CSP system through a heat exchanger. The hot tank operates at a temperature of 385ºC and the cold tank at a temperature of 292ºC. During the day the molten salt is pumped from the cold tank to the hot tank at a flow rate of 948 kg/s. The molten salt passes the oil-to-salt-heat exchanger, is heated up and continues as hot molten salt to the hot storage tank where the solar thermal energy is stored. At the same time the thermal oil is cooled down, before it returns to the solar field for reheating. Whenever heat is needed, the hot molten salt is pumped back to the cold tank, transferring thermal energy to the Rankine cycle, and cooling down the storage material while this is returning till the cold tank (Knutsson, 2008).

53

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

The molten salt media used in the Andasol 1 storage concept is HITEC, see section 6.1.1. The mixture provides better overall thermal characteristics than the salts individually, though closer to the nitrate single salt values. The downside of molten salt in CSP system is that of preventing the molten salt to freeze and solidify. There are ways to avoid solidification by heating through electric, internal resistance or impedance heating. Internal resistance heating is possible but cumbersome. Impedance heating gives a uniform heating but is as for electricity heating, electricity demanding. Care has to be taken of the choice of joints, seals and packings for molten salt components – graphite seals are incompatible with molten salt. Boron nitride is chemically suitable and could be an alternative (Moreno, 2003). Using Therminol™VP-1 or a similar synthetic heat transfer oil in the solar parabolic trough absorbers generally require an indirect system for tank storage concepts. High- temperature heat transfer fluids such as the thermal oils have relatively high vapor pressures at required operating temperatures. As a result storage tanks require pressurization which dramatically increases storage tank complexity and cost. Molten salts are commonly considered as the indirect storage medium. They offer the advantage of having a much lower vapor pressure but they have the disadvantage of high freezing temperatures (100ºC-220ºC), currently available used both as a storage medium and as a solar field fluid (McMahan, 2006).

A thermocline storage tank refers to the stratification that occurs in a tank with a large volume. A sharp fluid temperature gradient (stratification) develops between high and low temperature fluid caused by fluid buoyancy resulting from the variation in fluid density with temperature. The stratified thermal energy storage concept can be improved by adding a solid storage medium to the storage tank (McMahan, 2006). The purpose of the solid medium is to maintain thermal mass while reducing the required fluid volume for a given storage capacity. A temperature gradient forms in the solid as well as the liquid. The advantage is the use of inexpensive solid media (rock, sand, concrete) together with more expensive heat transfer fluids. Parasitic energy consumption and pressure drop could however be high is such system (Pilkington Solar Inc., 2000).

54

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Packed bed such as concrete storage is essentially the same as the stratified storage tank with added filler material. A solid concrete storage can also be termed high temperature sensible heat storage. The heat transfer fluid flow from the solar collector field is divided into smaller identical tubes, located in a square formation embedded in the solid block storage unit. The numerous tubes improve heat exchange between the heat transfer fluid and the storage as the area of heat transfer is larger. This alternative eliminates the heat exchanger otherwise necessary for an indirect molten salt rock bed storage system and leads to improved storage system performance and reduces cost (McMahan, 2006). However, it demands a structured rather than randomly packed bed, and the long-term structural stability of the cast concrete structures following multiple thermal cycles has to be evaluated. A thermocline solid storage tank of quarzite rock and silica sand offer an inexpensive filler material where quartzite has a higher thermal conductivity than most other rocks and does not deteriorate from prolonged exposure to the molten salt environment , making it attractive for solid TES (Brosseau, et al., 2005).

Figure 31. PCM-cascade TES concept for parabolic trough CSP (reproduced from Knutsson, 2008).

A phase change cascade storage tank is an example of storing latent heat, using several different salts with thermal properties and phase change temperatures in the range of 306-380 ºC. PCM cascade storage concept is shown in Figure 31, where

55

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

five tubular heat exchanger stages holding five salts made up of either single salts or eutectic mixture of several salts (see Table 4) covering the range of 306-380˚C.

Table 4. PCM cascade salts used in TES for parabolic trough CSP. (Knutsson, 2008))

Tubular heat Enthalpy exchanger PCM Average melting Average change Temperature density during melting # [ºC] [Kg/m3] [Kj/kg]

1 MgCl2/KCl/NaCl 380 1800 400 2 KOH 360 2040 134

3 KNO3 335 2109 95 4 KNO3/KCL 320 2100 74

5 NaNO3 306 2261 172

The cascade configuration salts absorb most of the heat as they melt at practically constant temperature, starting at 380ºC (1st salt), while the heat transfer oil temperature drops from 390 ºC at storage inlet to 300 ºC at outlet of the storage, where the last salt has the melting temperature of 306 ºC. The PCM cascade has storage density higher than sensible storage of equivalent capacity (McMahan, 2006).

Applications with packed beds or shell-and-tube vessels should consider the heat transfer limitation due to the pinch-point, the relatively small temperature differences between the PCM and the charging or discharging. At these points, a large heat transfer surface area is needed as the relatively small temperature differences drive the heat transfer.

The pinch-point characteristic is not an issue using direct steam generation but with cascade arrangement the pinch-point problem would be minimized by replacing it with a hypereutectic multi component salt for which the liquids have the same temperature as the maximum heat transfer fluid temperature, and the eutectic temperature is the same as the turbine inlet temperature. However such operation would require concurrent flow of the heat transfer fluid during the storage and utilization cycles (Luz Int., et al., 1988).

56

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

For the packed bed arrangement of PCM's, encapsulation and pressurized designs are the critical issues. Encapsulation refers to a protective outer layer which isolates the PCM from the heat transfer fluid and reduces their relative incompatibilities. Encapsulation for the packed bed arrangement should provide a high surface area for heat transfer and low resistance for heat conduction within the encapsulated sphere. A separate pressure vessel is however most likely necessary for collection of the heat transfer fluid off-gassing. One advantage of the shell-and-tube configuration over the packed bed configuration is that if the heat transfer fluid is contained in the tubes, the containment of the off-gases will be simpler and less expensive.

7 POWER BLOCK FEATURES OF CSP PLANTS

Figure 32. Example of a Solar thermal steam power plant for electricity production with no TES and fuel back-up. Figure 32 shows an example of a concentrating solar steam power plant where the solar thermal energy is delivered to the Rankine cycle to preheat and evaporate feedwater and to superheat steam. Superheating the steam will lead to an expansion mainly in the superheated region of the Mollier chart achieving a higher enthalpy drop during the expansion, leading to a higher power output. Only the last LP stages will experience expansion in the two phase region lowering the moisture losses in the turbine. Moisture contains water droplets that have an impact on the rotor, lowering the efficiency as well as causing erosion of the blades. Power production by

57

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

expansion of steam drives the turbine rotor i.e. turning the turbine shaft connected to an electrical generator. The turbine exhaust steam is condensed and returned to the Rankine cycle as feedwater. A deaerator both removes incondensable gases from the water/steam cycle and work as a water and heat storage, preventing system fluctuations. The high content of non condensable gases, especially oxygen could cause corrosion in the tube bank. The condensate is sometimes sprayed into the deaerator or it is arranged to pour down from a number of trays to reduce the ratio of non condensable gases in the feedwater. When heated steam is fed into the deaerator, the water droplets are heated to the saturation temperature, the incondensable gases is released and evacuated at the top of the deaerator.

7.1 Steam turbine

The steam turbine operates with a rotational shaft speed range between 1500-6000 rpm. For machines in electrical production the connection towards the electric generator have to be synchronized with the grid. For a 50 Hz grid the synchronous rotational speed is 1500 rpm. Both high pressure (HP) and low pressure (LP) turbines can have higher rotational speed than 1500, but have to be geared down before connecting to the electric generator. Also the HP turbine and the LP turbine can be connected to the same electric generator but with different synchronous rotational speed. For example an HP turbine rotating at 12000 rpm and a LP turbine rotating at 5000 rpm can both be connected to a four pole generator at 1500 rpm through gears.

For steam turbines in CSP-plant some characteristics are interesting to distinguish: saturated vs. superheated steam; reheat and regenerative cycles; and condensing, backpressure and extraction turbines.

The live steam to the steam turbine is preferably superheated by a steam heating tube bank in the heat exchanger unit. In some cases the live steam is taken directly from the evaporator stage leading saturated steam to the turbine. The difference between superheated and saturated steam is in superheated steam conditions the saturated steam is heated at the same pressure up to a higher temperature which gives an increased enthalpy drop at expansion and a higher

58

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Rankine cycle efficiency. For the saturated steam case, the moisture content is reduced before the steam is let into the turbine. In more expensive turbine plants the HP and the LP blading is separated into two or more separate casings. The exhaust steam from the HP is reheated before entering the LP turbine to conserve the steam quality. For reheat turbines it is common with three separate casings dividing the turbine into; high-, intermediate (IP) and low pressure turbines. The IP expands the steam from the hot reheat HP outlet steam to the inlet LP element. A regenerative cycle uses extracted steam from intermediate pressure outlets to preheat the feedwater to the steam generator. When the steam plant producing electricity and no process heat, the best efficiency is reached when the steam turbine operates with low back-pressure at the outlet before the exhaust steam condenses. When using the steam for process heat, for example district heating, drying, boiler etc., the back- pressure has to be higher which allow higher outlet steam temperature to be used for process heating. In the electricity production condensing turbines are used where the back-pressure is low, 0.03 - 0,06 bar (Widell, 1965; Ecostar, 2005) and the enthalpy fall in the turbine is transferred to maximum mechanical power. In Rankine cycles with a deficit of low temperature heat, steam is extracted from different points in the turbine to heat feed water in pre-heaters to raise the temperature before the feed-water is evaporated by the high temperature source. This is to reduce exergy losses in the cycle and thereby increase the cycle efficiency (low pressure preheater in Figure 32).

There are significant differences when using the steam turbine in a CSP plant as compared to other applications. Fluctuation of temperatures and loads are larger and therefore should such steam turbines for example be flexible and not too sensitive to thermal stresses. The operation modes for CSP plants differ from a regular steam plant that operates around the clock all year, only stopping the turbine for service, inspection and maintenance and the start and stop events are more frequent for CSP plants than for traditional applications. Low-load operation like gland sealing and rotor balancing, along with consequences due to heat losses take on much more importance when steam turbines are stopped and restarted daily.

59

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Gland Sealing Steam gland sealing can be used for three reasons: to maintain a vacuum in the turbine; to prevent oxidation of the turbine via air entrainment and to minimizing heat losses during stand still (if it is economical beneficial to produce auxiliary steam). Most steam turbines exhaust into a vacuum where a multiple number of seals dissipate the turbine pressure and the velocity to the ambient surroundings. Gland sealing is used where the turbine shaft passes through the casing and shaft walls.

Gland steam 1.1 – 1.3 bar

Air Turbine Atmospheric Sealing Sealing Sealing pressure

1 bar

Leakage system

0.98 bar

Figure 33. Example of Gland Sealing steam system. Viewing Figure 33, the gland steam supply operates with a pressure of 0,1 – 0,3 bar. Gland steam above atmospheric pressure is introduced and flows inward to be condensed in the turbine and outward as a slight positive flow to prevent atmospheric air from entering, exiting into the leakage system at pressure under atmospheric. The gland steam flows inwards to the turbine when auxiliary heat is needed at stand still and at start-up. Any air leaking in is forced into the leakage system by its lower pressure. The air/steam that flows into the leakage system is condensed and reused in the steam cycle. The amount of gland steam needed depends on several input parameters, where pressure drop in the system is the most important parameter.

When the turbine is shut down the pressure decreases in the system and the turbine starts to cool down. The small pressure difference between the ambient air and the leakage system could lead to air/steam leakage into the turbine, if the turbine is cooled down to a level where a lower pressure is induced in the turbine. Cold starts are to be avoided as excessive condensation creates too high moisture content

60

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

followed by, among other problems, vibrations of the rotor blades. Also the large temperature difference between cold start and full operation leads to thermal stresses and risk of material expansion and stresses in rotor and casing followed by changes in geometry.

Turbine stresses Elliot (1998) provides details about material aspects of the steam turbine, which has to be able to withstand material stresses due to high pressure loads, high temperatures and temperature transients at start and stop phases. There are numerous temperature zones inside a turbine casing leading to temperature gradients in both axial, radial and tangential direction (cylinder) across the casing wall. Other factors that relate to steam turbine thermodynamics are cycle variables as for reheating, initial steam conditions, feedwater preheating, exhaust steam conditions, blade efficiency, flow losses etc. These constraints could lead to tensile or creep rupture-failure, repeated plastic strain leading to fatigue damage and cracking. The choice of material has to consider all the criteria. If the heat transfer is high, the casing material thickness is large and the thermal diffusivity is low, the thermal gradient will be large across the casing wall. On the other hand if the heat transfer is high, the casing material thickness is small (leaner construction) and the thermal diffusivity is high, the faster the turbine will reach a uniform temperature, and the thermal gradients will be lower across the casing wall, extending the life-time of the turbine.

The material of a steam turbine casing are of ferritic steels that has higher thermal conductivity and lower thermal expansion which is preferred in pressure-vessel materials up to 566 C. High-temperature rotors, capable of operating at temperatures above 482 ˚C use Cr-Mo-V composition 12 % Cr alloy. Cr-Mo-V has higher oxidation resistance while 12 % Cr alloy has higher strength at high temperatures. LP rotors use a Ni-Cr-Mo-V alloy that has a low transition temperature but is sensitive to temperature embrittlement above 371 ˚C. Below 343 ˚C Hp rotors can use LP rotor material as well as carbon steel.

61

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

7.2 Dynamic Operation of Steam Turbine in CSP Plants

The variation in thermal energy supply due to fluctuation in insolation is the main difference between solar power and fossil power, that supplies thermal energy continuously. In Figure 34 it is evident that not only varies the insolation by month and season but also from year to year. A trend curve shows the spring and summer to have the highest annual insolation of 175 – 260 Wh/m2 at a particular site. Solar power technology meets the challenge to levelize the differences and achieve a stable electricity production.

Figure 34. Annual direct Insolation period 2000 – 2008. (source: Abengoa Solar)

Diurnal operation of the CSP plant can be executed in several different ways depending on the CSP concept, TES-system and electrical dispatch demand. In Figure 35 the chosen diurnal CSP operation is to deliver thermal energy both to the TES and the turbine at the same time. When the solar insolation becomes insufficient, the TES will prolong the electricity production several hours, resulting in a decrease in plant efficiency, hence the dip of the turbine curve (violet curve) and a slight dip in the electricity delivery (red curve). Interesting is the curve of dumped thermal energy where the TES storage is not charged with the thermal energy delivered by the solar collector field that is not used by the Rankine cycle. The reason could be an oversized collector field.

62

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 35. Diurnal operation of a CSP plant – TES storage, prolonged electricity production. (source: FLAGSOL)

Another operational concept is to meet the peak demand of electricity in the afternoon (Figure 36) where the TES is charged before production of electricity.

Figure 36. Diurnal operation of CSP - plant. TES storage as intermediate stage. (source Solar Millenium) When fully charged both the collector field and TES deliver thermal energy to the Rankine system. The storage is also charged a time after electricity production stop. With simultaneously discharging of TES and Solar field, the time of dispatchable electricity is less compared to the electricity production in Figure 35. The TES storage plays the role of an intermediate stage in delivering the thermal energy to the Rankine cycle.

63

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

The following graph shows the behaviors of a 5MWe DSG plant higher during winter operation. As the DGS solar power plant is more sensible to the insolation the diurnal operation is highly influenced by the azimuth tracking. In winter as the parabolic troughs follow the sun’s movement over the hemisphere at noon, the collector field risks not to be able to absorb the thermal energy, as the control of the troughs are designed by summer azimuth. Hence the dip at noon visualized by the green curve in Figure 37, followed by the steam flow (red curve) and electricity production (dotted curve).

Figure 37. Diurnal operation DSG CSP-plant for month of January (Arza, et al., 2004).

The design of the solar parabolic trough field is based on winter climate conditions, i.e. low insolation. The design oversizes the collector field for summer conditions, (Figure 38) which increases the absorbed thermal energy to a level where it is necessary to deflect the troughs, giving the curve of summer a “flat line” at maximum DNI of the green curve – maximum absorbed thermal energy has been reached and the temperature levels are too high in the solar collector field. By deflecting the solar reflectors over temperature can be avoided.

64

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 38. Diurnal operation DSG CSP-plant for month of June (Arza, et al., 2004).

Figure 39 show the temperature, pressure and power of a CSP steam turbine in a diurnal operation mode. Starting with the stop process at the end of the CSP plant operation, the pressure followed by the power decrease rapidly; the inlet temperature still high, slowly reducing as high temperature steam is no longer fed to the turbine. Next phase, the turbine slowly cools down during stand-still phase until the point where the turbine is in start phase where the inlet pressure and temperature is increasing until the turbine can deliver mechanical work for electrical production.

65

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 39. Steam turbine diurnal operation in CSP. (source: Siemens Industrial Turbomachinery AB)

By diminishing the cool down of the steam turbine during night stand still, thermal gradients during the following start up can be reduced. This allows for faster turbine start up and linger life time for thermally exposed turbine components.

66

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

7.3 Organic Rankine Cycle Plant

The most common uses for ORC are for geothermal, solar power generation, waste heat recovery and remote power. The ORC solar power plant technology is more compact, shows higher versatility and simplicity compared to traditional Rankine cycle powerplants and are usually projected for <5 MWe. (McMahan, 2006) The design of low or no-maintenance power cycles is desirable in remote and low-output applications where no maintenance personnel are available.

An ORC solar power plant operates at lower temperatures, lower pressures and condenses at above-atmospheric pressures. The solar collector operating temperature range is between 300ºC - 400ºC on the hot side and 15 ºC - 40 ºC on the cold side (McMahan, 2006). This means that an inexpensive heat transfer fluid may be used. Organic working fluids require smaller condensing equipment minimizing size and complexity of the system.

ORC –cycle fluid thermal characteristics are thermal stability, toxicity, flammability and cost. As an example, toluene has good thermal characteristics regarding stability, temperature range and high energy density but condenses at sub- atmospheric pressures. The n-pentane on the other hand operates at super- atmospheric pressures over a wide range of condensing conditions (McMahan, 2006).

Most organic fluids show, to varying degrees, drying behavior resulting in a superheated vapor at isentropic expansion16. See Figure 40 for a schematic ORC showing key components. As the expansion in an organic working fluid normally is shorter compared with steam expansion mainly due to a condenser pressure around or above atmospheric pressure the available energy in the working fluid stream leaving the turbine exhaust is high. This energy can be converted into work by a second expansion in an organic fluid with a lower condensing temperature.

16 The end expansion in the superheated area. 67

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

ViewingFigure 40, the ORC's working principle is as follows: heat transfer fluid circulates through a parabolic trough solar field, transferring its heat energy to the high temperature working fluid through the evaporator [1] and returning to an accumulator before being pumped to the solar field to recirculate. The high temperature fluid vapor enters the high temperature turbine expander and produces mechanical work for electricity generation. The turbine exhaust steam is condensed in a recuperator [2], preheating the ORC fluid prior to its entry into the evaporater [1]. The high temperature fluid condenses further in a condenser/evaporator [3], where it is ‘cascaded’ down evaporating a low temperature fluid. The high temperature fluid returns to an accumulator [4] to recirculate. Additional power can be produced in a bottoming cycle with low-temperature fluid (Prabhu, 2005), although not all ORC concepts utilize this ([5] trough [7]).

The drying behavior that results in a superheated turbine exhaust simplifies cycle design at low temperature. The turbines in ORC often involves a single stage driven by higher fluid density at typical turbine operating conditions and smaller expansion pressure ratios that gives high condensing pressures. The highest ORC efficiency cycle utilized turbine reheat and recuperator. According to two studies, the ORC efficiency could reach 23 - 24 % with best design (McMahan, 2008; Price, et al., 2002). Another study showed that a tier-cascade design with reheat and recuperator yielded a net efficiency of about 24 % with a heat transfer fluid temperature of 287 ˚C and about 30 % net efficiency, when heat transfer fluid temperature from the solar collector field was increased to 462 ˚C (Prabhu, 2005) (It should be noted that the former design included an air cooled condenser and the latter a water cooled condenser). The design of reheating, recuperator and tier-cascade for higher efficiency has an additional turbine and heat exchanger, making the system more complex. Feedwater heating, usually applied in Rankine cycle, has no important improvement on ORC performance (McMahan, 2008).

68

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Figure 40. ORC Tier-Cascade/Recuperator/Reheat steam turbine cycle with high- and low

working fluid, giving the best net efficiency. 5MW.

Figure 41. Schematic Diagram of Saguaro Solar Trough Plant Operation.

The 1-MW APS Saguaro Solar Power Plant 8, (Figure 41), combines solar trough technology with an ORC power block. Its size was a factor to employ ORC technology rather than a conventional steam cycle that would be more efficient for a larger scale plant. The Saguaro solar power plant features more than 10 km2 of parabolic mirrors aligned in six rows, tracking the sun on a single axis and

69

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

concentrating the solar energy 60 times, generating with a generation capacity of about 6 MW in summer mode of 6843 kW. The heat transfer fluid is mineral oil, heated to approximately 300°C. The oil passes through a heat exchanger to vaporize an n-pentane working fluid, which drive a turbine and is then captured, compressed back to a liquid state, and recirculated to be used again. The plant has a six hours thermal energy storage. (Renewable energy world, 2009) More details can be found in Appendix 3, Table 13 and 14.

7.4 Cooling Water Requirements

In a Rankine steam solar power plant water often is used in the steam cycle condensate process, cooling for the condenser, and washing mirrors. With wet cooling, the cooling tower represents approximately 90% of the raw water consumption; the steam cycle represents approximately 8% of raw water consumption, and mirror washing represents the remaining 2% (EPRI, 2006). Soiling- resistant glass is being explored to further reduce the mirror washing requirement. Steam power plants driven by trough and tower systems can utilize dry cooling technology, in order to significantly reduce water consumption but it comes with a decrease in turbine efficiency and an increase in electricity cost due to, among other factors, higher condenser temperatures.

Table 5. Water usage of different plants based on data from US Department of Energy, (NREL 2006).

Total water usage of Plant type power plant output [m3/MWh] Stand‐alone steam plant 2,84 Stand‐alone steam plant with dry cooling 0,11 Parabolic Trough with wet cooling 3,78 Parabolic trough with dry cooling 0,30

Table 5 shows the large difference in water usage between wet and dry cooling. A decrease of 92 - 94 % water usage can be achieved by choosing dry cooling. Consideration has to be taken to the change in efficiency as wet cooling has proven to give higher plant efficiency as air cooled systems generates a higher condenser pressure generating a lower steam cycle efficiency

70

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

8 CONLUDING REMARKS

The technology of CSP has been known for centuries but has not been effective enough to be commercialized on a large scale until now. During first decade of the 21st century the concentrating solar power around the world has seen a rapid increase of interest from governments and industry as well as from other groups than the environmental organizations. New areas of expertise are developing and work opportunities are increasing, promising a sustainable labor market ahead. The commercializing of the CSP technology will give incitements to solutions and improvements of the challenges that are faced by the market today including solar collectors and control systems, storage solutions and its media to more flexible and durable power block integration. Quite large diurnal fluctuations demand fast response from electricity generation an extensive transmission grid to help stabilizing and maintaining the balance between production and demand. Technology expertise needed, comprises of for example mechanical engineering, material and computer science, system implementation and chemical and environmental field of knowledge.

9 TABLE OF SELECTED CURRENT AND PROJECTED CSTEPP

A large number of CSP projects are either in operation or under way to be operational during the next 10 years. The CSP projects include both stand-alone concepts as well as combine cycle plants. The combine cycle plants are the most common application, as solar thermal is used as bottoming cycle of the electricity production, meeting the peak power demand and reducing the fossil fuel consumption. Many such projects are in operation and under way in the countries around the Mediterrean sea as well as elsewhere in the world. The report considers mostly the stand-alone concept and a selected number can be seen in. The table does not implicate a full list of all projects but rather showing the raising interest in these kinds of power production technology and the increasing plant availability.

71

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Table 6 a) Selected CSTEPP projects, operational and planned plants.

Power Size Thermal [MW] Projectname Site Company CSP‐type Thermal Fluid Storage Turbine Type [MWe] 280 Solana (US, Az) Arizona Abengoa Solar Parabolic Trough Thermal oil Molten salt NA NA 50*4 Andasol 1‐4 Spain Abengoa Solar Parabolic trough Thermal oil Molten salt Dual casing geared Reheat Siemens 50

AS/Cobra Biphenyl/Diphenyl 50*3 Extresol 1‐3 Spain Group Parabolic Trough Oxide Molten Salt Dual casing geared Reheat Siemens 50

AS/Cobra Biphenyl/Diphenyl 50*2 Manchasol 1‐2 Esp Group Parabolic Trough Oxide Molten Salt Dual casing geared Reheat Siemens 50 50*3 Solnova 1‐3 Spain Abengoa Solar Parabolic Trough Thermal oil NA Siemens 50 Diphenyl Diphenyl 50 Vallesol Spain Torresol Parabolic Trough oxide Molten salt NA 50

64 Nevada Solar 1 Nevada Acciona Parabolic Trough Therminol VP-1 NA (o,5 hr) Dual casing geared Reheat Siemens 75 NA SOLENHA Haut Alpes Solar Euromed Parabolic trough NA Molten Salt NA Saguaro Solar Food‐rade mineral 1 Power Plant Arizona Acciona Parabolic Trough oil NA n‐pentane Ormat Israel 1,16 Steel drum (1 11 PS10 Spain Abengoa Solar Power Tower DSG hr) Saturated Steam Siemens 11 Steel drum (1 20 PS20 Spain Abengoa Solar Power Tower DSG hr) Siemens 20

72

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Table 6 b).Selected CSTEPP projects, operational and planned plants.

Power Size Thermal [MW] Projectname Site Company CSP‐type Thermal Fluid Storage Turbine Type [MWe] 2*100 Ivanpah Solar Mojave Desert +200 Power Complex (US, Ca) Bright Source Power Tower DSG NA Dual Casing Reheat Siemens 123 Lancaster 5 Sierra SunTower (Ca, US) eSolar Power tower DSG NA GE 5 Power Tower, 2600 17 Solar Tres Andalùcia Sener Heliostats Molten Salt Molten Salt Two Cylinder Reheat Siemens 19 Dual Casing, double‐flow 177 Carri Energy Farm California (US) Ausra CLFR Saturated steam Steam drum down exhaust turbines NA 177 Kimberlina Solar Bakersfield 5 Thermal (US, Ca) Ausra LFR DSG NA NA 5 Novatec Solar 1,4 Puerto Errado 1 Murcia (Spain) Spain LFR DSG Thermocline NA 1,4 Solar Two ‐ Phase Imperial Valley, Stirling Energy 300‐900 1‐3 (US, Ca) System Dish engine NA NA NA NA Solar One ‐ Phase Mojave Desert, Stirling Energy 500‐800 1‐2 (US, Ca) System Dish engine NA NA NA NA 1,5 Maricopa Arizona (US) SES Dish engine NA NA NA NA

73

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

10 BIBLIOGRAPHY

Abengoa Solar NT current projects, 2008, “Current Projects Parabolic Trough Collector”. Retrieved from: http://www.abengoasolar.com/sites/solar/en/abengoa_solar_nt/current_projects/troug h/index.html, accessed 20090110.

Abengoa Solar, 2008, “Concentrating Solar Power, Parabolic Trough”. Retrieved from: http://www.abengoasolar.com/sites/solar/en/technologies/concentrated_solar_power/ parabolic_trough/index.html, accessed 20090110 f

Abengoa Solar, 2009a, ”Abengoa Solar, Solar Power for a sustainable world”, White paper on Tower technology, Retrieved from http://www.abengoasolar.com/sites/solar/resources/pdf/en/PS10.pdf, Accessed 20090129.

Abengoa Solar, 2009b, “Abengoa Solar, Solar Power for a sustainable world, Solution to Global Climate Change, Parabolic Trough Plants”, Solnova_1, Retrieved 20090110 from: http://www.abengoasolar.com/sites/solar/resources/pdf/en/Solnova_1.pdf

Andraka C.E., Diver R.B., Moreno J. B., Moss T.A., Adkins D.R., 1994, “Recent reflux receiver developments under the US DOE program”, Solar thermal Technology Department, Sandia Natioanl Laboratories, SAND94-1758C

ANU, 2008, “Closed loop thermochemical energy storage system using ammonia”, Solar thermal energy, Solar thermal group research, October 2008, Australia National University, Retrieved from http://solar- thermal.anu.edu.au/high_temp/thermochem/index.php, accessed 20090818, Page last updated: 26 October 2008.Thermal Energy Research

74

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

APP, 12 Juni 2002, “Compact linear Fresnel reflector (CLFR) power plant technology”, Applied Physics, University of Sydney, Australia,, Retrieved from http://www.physics.usyd.edu.au/app/research/solar/clfr.html, Accessed 20090818

Ausra, 2007, “An Introduction to Solar Thermal Electric Power”, Ausra, Inc.2585 East Bayshore Rd. Palo Alto, CA 94303,

Bockamp S. Dr., Griestop T., Fruth M., Dr. Ewert M., Lerchenmüller H-J., Mertins M., Morin G., Dr. Häberle A., Dr. Dersch J., 2003, ”Solar Thermal Power Generation”, E.ON Engineering GmbH, E.ON Energie AG, Fraunhofer Institute for Solar Energy Systems (ISE), PSE GmbH, German Aerospace Centre (DLR), Germany.

Brightsource, 2009, retrieved from http://www.brightsourceenergy.com/faq.htm,

Accessed 20090215

Brogren M., 2004, “Optical efficiency of Low-Concentrating Solar Energy Systems with Parabolic Reflectors”, Uppsala University, Faculty of Science and Technology, 934.

Brosseau D., 2003, “Testing Thermocline Filler Materials for Parabolic Trough Thermal Energy Storage”, Sandia National Laboratories, Thermal Storage Workshop for Parabolic Trough System, February 20, 2003.

Cordeiro, P., 1997, ”Start Mission to Brazil”, IEA Solar Power and chemical Energy Systems, SolarPaces, start report 2/98.

DiBella F.A., Gwiazda J., 2009, “ Pit Power Tower, A novel application of Pneumatic and Solar power Tower Technology ”, retrieved from http://www.altenergymag.com/emagazine.php?issue_number=09.02.01&article=gwia zda&id=1327, accessed 20090810

75

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

De Vos A., 2008, “Thermodynamics of Solar Energy Conversion”, Universitiet Gent, Belgium, Wiley-VCH Verlag GmbH &Co. KGaA, Weinhem. ISBN 978-3-527-40841-2

Diver, Richard B., et al. "The Advanced Dish Development System Project." Washington D.C., April 21-25.

Duffie J., A., Beckman W., 1991, “Solar Engineering of Thermal Processes”, John Wiley & Sons, Inc., ISBN 0-471-51056-4.

Earth 2 Tech, 2008, Retrieved from: http://earth2tech.com/2008/06/26/10-questions- for-esolars-ceo-asif-ansari/, Accessed 20090130.

Elliot C.T., Chen K., Swanenkamp R. C., 1998, “Standard handbook of Powerplant Engineering”, 2nd edition, The McGraw-Hill Companies Inc., United America, Printed by R.R Donnelley & Sons Company, ISBN 0-07-019435-1.

EPRI, 2006, “Solar Thermal Electric Technology: 2006”, Palo Alto, CA: 2007. 1012731, Technical Update, March 2007, EPRI Project Manager - T. Key.

EPRI , 2008, “New Mexico Central Station Solar Power Summary Report”, 2008 Electric Power Research Institute, Inc., Retrieved from: http://www.abengoasolar.com/sites/solar/resources/pdf/en/Solnova_1.pdf, accessed 20081210

ESolar,2008, Brochure “eSolar, Utility-Scale Solar Power” Retrieved from: http://www.esolar.com/solution.html, Accessed 20090130.

EC (European Commission), 2007, “Concentrating Solar Power, From research to implementation”, 2007, ISBN 978-92-79-05355-9, Europoean Communities.

Fairley P., 2008, ”Technology Review Solar without the Panels”, retrieved 20090110 from http://www.technologyreview.com/business/20356/page2/

Floyd-associates, 2009, “Solar Industry Report, A Trend towards Solar Thermal”, Retrieved from: http://www.floyd-associates.com/solar.pdf, Accessed 20090130.

76

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Fraser P. R., 2008, “Stirling Dish System Performance Prediction Model”, master of science (mechanical engineering), University of wisconsin-madison, US, January 14, 2008

Gil, A., Medrano, M. and Cabeza, 2008, “State of the art of high temperature storage in thermosolar plants” L. F. University of Lleida, GREA Innovació Concurrent, Pere de Cabrera s/n 25001 Lleida, Spain

Glatzmaier G., Blake D., Kang J., 2008, “Thermal Storage - Advanced Fluid Development”, NREL, 2008 Solar Annual Review Meeting

GMI, 2004, Global Market Initiative, “The Concetrating Solar Power”, Uniq Werbeagentur.

Goswami Y., Dr., Kreith F., 2008, “Energy Conversion”, p.4, Published by Taylor & Francis, /th of Juky 2008, Boca Raton.

Luz International Ltd., Solar Energy Research lnstitute, CBY Associates Inc., 1988, SERI monitors Luz Int., , Gupta B., 1988, “Phase-Change Thermal Energy Storage”, Symposium October 19-20,1988, Helendale, California, U.S. Department of Energy, Contract No. DE-AC02-83CH10093

Hoshi A., Mills D.R., Bittar A., Saitoh T. S., 2004, “Screening of high melting point phase change materials (PCM) in solar thermal concentrating technology based on CLFR”, Ichinoseki National College of Technology, Japan, University of Sydney, Tohoku University, Japan, 30 December 2004

IEA, 1998, International Energy Agency, “Solar Power and Chemical Energy Systems”, Solar Paces START Mission to Brazil Report 2/98, Final Report by Patricia Cordeiro, May 5-9 1998

77

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Kearney D. W. & Associates, Flabeg Solar International, KJC Operating Co., Nextant (Bechtel), NREL,Sandia Natl. Lab, MWE, 2001, “Engineering Evaluation of a Molten Salt heat transfer fluid in a Parabolic Trough Solar Field”, NREL Contract No. NAA-1-30441-04, Presentation .ppt

Lovegrove K., Zawadski A., Coventy J., 2006, “Taking the ANU Big Dish to commercialization 2006”, Department of Engineering, Australian National University

Mancini, T., Heller, P., Butler B., et al., 2003, "Dish-Stirling Systems: An Overview of Development and Status." Journal of Solar Energy Engineering 125 (May 2003). Sandia National Laboratory, DLR

McMahan A. C., 2006, “Design & Optimization of Organic Rankine Cycle Solar- Thermal Powerplants”, University of Wisconsin-Madison, US.

Mills D., Morrison G.L., Le Lièvre P., 2004, “Design of a 240 MWe Solar Thermal Power Plant”, Eurosun 2004 Conference, School of Physics University of Sydney School of Mechanical and Manufacturing Engineering, University of New South Wales, Australia,

Moens L., 2004, “Advanced Thermal Storage Fluids for Solar Parabolic Trough Systems”, GCEP Solar Energy Workshop, SunLAb, Stanford University, Palo Alto (US, CA)

Moreno J.,2003,” 2002 Evaluation of Interconnections with Molten Salt heat transfer fluid in a Trough”, Sandia National Laboratories, SF2002, NREL Trough Thermal Storage Workshop Golden, February 20-21, 2003)

Nava P., Herrmann U., 2006, “Thermal Storage Concept for a 50 MW Trough Power Plant in Spain”, Trough Workshop in Lake Tahoe, February 13, 2006, Incline Village

NREL, 2006, “Parabolic Trough Work shop, Cooling for Parabolic Trough Power Plants”, Incline Village, Nevada, February 14.

78

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Ortega(1) I. J., Burgaleta I. J., Téllez F. M., 2009, “central receiver system (crs) solar power plant using molten salt as heat transfer fluid”, Retrieved from: http://www.sener.es/eportal_docs/general/file- cwa0fcc36424ab41b7bf04/solartres.pdf, accessed 20090131.

Pilkington Solar International Gmbh., 2000, “Survey of Thermal storage for Parabolic Trough Power Plants”, NREL, September 13, 1999.June 12, 2000, Cologne, Germany, September 2000 . NREL/SR-550-27925.

Pitz-Paal R., Dersch J., Milow B., 2005, “European Concentrated Solar Thermal Road Mapping”, Ecostar, SES6-CT-2003-502578. DLR Prabhu E., 2005, “ Solar Trough Organic Rankine Electricity System (STORES) Stage 1: Power Plant Optimization and Economics “, Reflective Energies Mission Viejo, California

Price H., Hassani V., 2002, “Modular Trough Power Plant Cycle and Systems Analysis”, January 2002 • NREL/TP-550-31240

PSA, 2006, “Annual Report 2006”, Plataforma Solar de Alméira, Retrieved from: http://www.psa.es/webeng/techrep/techent.html, Accessed 20090131

Renewable Energy World, 2006“Saguaro Solar Power Plant”, 2006, retrieved from: http://www.renewableenergyworld.com/rea/news/article/2006/04/solar-energy- milestone-reached-in-the-arizona-desert-44696, accessed 20090514.

Saguaro Solar Power Plant, 2006, “SP017, Saguaro Solar Trough Power Plant”, July 7 2006.

Sargent & Lundy LLC Consulting Group, 2003, “Assessment of Parabolic Trough and Power Tower Solar Technology Cost and Performance Forecasts”, Chicago, Illinois, NREL, October 2003, NREL/SR-550-34440.

79

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Solar Millennium, 2009 “Andasol, The World’s Largest Solar Thermal Power Plant Project Development in Andalucia” (Spain), Retrieved 20090130.

Solar Reserve, 2009, Retrieved from: http://www.solar-reserve.com/FAQS.html & http://www.solar-reserve.com/technology.html, Accessed 20090215

Solúcar Solar S.A, 2006, “Final Technical progress report, PS10, 10MW Solar

Thermal Power for southern Spain”, November 2006, Project NNE5-1999-356.

Stirling Energy, 2008, ”Suncatcher”, Retrieved from: http://www.stirlingenergy.com/technology/suncatcher.asp, Accessed 20081213

Tamme R., 2003,“Workshop on Thermal Storage for Trough Power Systems”, DLR, February 20-21 2003 , Golden CO, USA

Tyner C., Kolb G., Prairie M. (Sandia National Laboratories), Weinrebe G., Valverde A. (DLR, PSA) Sanchez M. (EU), 1996, “Solar Power Tower Development: Recent Experiences”

Widell T., Professor, 1965, ”ÅNGTEKNIK, Kraftprocesser”, Institutionen Ångteknik, KTH, Stockholm.

Zalba B., Marın J.Ma, Cabeza L. F., “Review on thermal energy storage with phase change: materials, heat transfer analysis and applications”, Universidad de Zaragoza, Universit_aria Polit_ecnica, Universitat de Lleida, Spain, ZAE Bayern Germany, 2002 Elsevier.

Zarza E., Rojas Ma E. Gonza´lez L, Rueda J.Ma, C., F., 2005, “INDITEP: The first pre-commercial DSG solar power plant”, Plataforma Solar de Almerı´a, CIEMAT, INITEC Energı´a,, Iberdrola Ingenierı´a consultorı, 2005 Elsevier.

80

Overview of Concentrating Solar Power for Electricity Production, with Emphasis on Steam Turbine Aspects

Zarza E., Rojas Ma E. Gonza´lez L, Rueda J.Ma, C., F., 2006, “Saturated steam process with direct steam generating parabolic troughs”, Plataforma Solar de Almerı´a, CIEMAT, INITEC Energı´a, Iberdrola Ingenierı´a consultorı, 2006 Elsevier.

81

APPENDIX

1. Concentrating ratio – C

The way of achieving the maximum concentrating ratio on earth is to create an image of the sun, i.e. the solar dish at one focal point and its image at the other focal point on earth’s surface taking into account the dilution of the solar flux density1. The solar 2 radiation leaves the surface area of the sun of 4π(Rs ) and when reaching the earth surface the same amount of solar radiation distributed over a larger surface of 4πr2 (see Figure 1). This dilution ratio becomes the solar radius to the sun-earth distance. From a “collector” point of view the dilution ratio is the solar irradiation received by the collector from a solid angle [ω] to the total hemisphere, i.e. the aperture factor [f]. (De Vos, 2008)

Figure 1. The solar point of view, the dilution of solar irradiation flux and the aperture factor from a solar collector point of view, where it sees the solid angle of the Hemisphere.

Eq. 1

1 4 Stefan-Boltzmann flux σTs . 82

6 The Rs is the sun’s equatorial radius of 696 x 10 m and r is the sun-earth average distance of 150 x 109 m2. f= 21,6 x 10-6

The maximum concentrating ratio is limited by physical laws as the solar rays cannot be focused into an infinitely small area, therefore the maximum solar power incident on the solar collector should be equal to the solar flux density at the sun’s surface [ , with correction to the dilution [f] of the same and the solar flux concentrating ratio [Cg] of the solar collector (De Vos, 2008):

Eq. 2

Where the solar surface temperature Ts is 5762 K, and Stefan-Boltzmann constant σ equal to 5,67x10-8 W/m2K4. This leads to the theoretical maximum value of the solar flux concentrator factor (when the solar absorber “sees” the solar disc in a 2π solid angle) of Cg = 1/f , i.e Cgmax = 46 296 ≈ 46 300. The theoretical value of 46 300 is the upper limit for the point focus system reciever.

A linear system has a different geometrical shape and “sees” a different size of the hemisphere than the point focus system. It does not track the sun optimally but tracking the sun’s movement single axis over the hemisphere north/south direction or east/west direction. The maximum solar flux concentration factor for a linear focus system becomes (Duffie J.A., et. al., 1991):

Eq. 3

The concentrating ratio of a solar collector is based on geometrically figures of the collector’s aparture area and absorber area, the so called geometrical concentrating ratio [Cg]. The aperture area is the surface into which the solar flux entering the collector, i.e. reflector and the reciever area is the area that absorbs the solar energy.

2 Average value cause to the earth’s elliptic path around the sun; at perihelion r=147 Gm, at aphelion r=152 GM. 83

Aparture Area

Sun

Reciever

Figure 2. Solar irradiation onto a collector’s aperture and reciever.

Eq. 4

Cg should be equal or larger than one. The concentrating ratio is direct proportional to a decrease in solar flux density and therefore to a larger absorber area of the reciever and higher thermal losses. The aim for improvements is to increase solar flux density and decreasing the heat losses.

84

2. CSP Power Plants

ɳT - Thermal efficiency (Solar thermal cycle)

ɳsteam cycle - Rankine cycle efficiency (Steam cycle)

ɳsolar to electric – CSP plant efficiency (Solar thermal to electric production)

Table 1. Solnova-plant, parabolic trough CSP, Seville, Spain.

Parabolic Trough Design Data Plant ηT Thermal Fluid # loops troughs/loop [%] [ᵒC] [m2] 90 50 MWe 57 Synthetic oil 4 400 300 000

Table 2. Solnova-plant, parabolic trough CSP Power cycle data, Seville, Spain.

Power cycle ηsteam cycle Turbine ηSolar to electricicty Power Generated CO2-savings [%] [˚C/bar] [%] [GWh/year] [ton/year] Rankine 36 Steam 19 114,6 31 200 370/100

Table 3. AndaSol 1-3, parabolic trough CSP, plant data, Spain.

Parabolic Trough Design Data ηT Thermal Fluid Storage # loops troughs/loop Plant [%] [ᵒC] [ᵒC] [m2] [Hrs] 90 50 MWe NA Synthetic oil Molten salt 4 400 385 2 000 7,5

Table 4. AndaSol 1-3, parabolic trough CSP, Power cycle data, Spain.

Power cycle ηsteam cycle Turbine ηSolar toelectricicty Power Generated CO2-savings [%] [%] [GWh/year] [ton/year] Rankine NA Steam NA 180 48003

3 Authors calculation: 450 kWh/12 t CO2 85

Table 5. DISS/INDITEP Superheated steam DSG parabolic trough CSP data.

Parabolic Trough Design Data ηT Thermal Fluid # loops troughs/loop Plant [%] [ᵒC] 2 [m ] [MWe] [bar] 7 5 65,1 Superheated steam (x=0,85) 10 410 (8 evaporation 2 superheating) 3971 70

Table 6. DISS/INDITEP Superheated steam DSG parabolic trough CSP, power cycle data.

Power cycle ηsteam cycle ηSolar to electricicty [%] [%] Rankine 25,9 16,4

Figure 3. CSP plant for Direct superheated steam generation.

86

Table 7. DISS/INDITEP Saturated steam DSG parabolic trough solar plant.

Parabolic Trough Design Data ηT Thermal Fluid # loops troughs/loop Plant [%] Turbine 2 [m ] [MWe] [ᵒC] [bar] 9 5 66,9 Saturated steam 8 280 5106 65

Table 8. DISS/INDITEP Saturated steam DSG parabolic trough solar plant.

Power cycle ηsteam cycle ηSolar to electricicty [%] [%] Rankine 24,9 16,2

Table 9. PS10-plant, Power Tower CSP, Seville, Spain.

Heliostat Tower Design Data ηT Thermal Thermal CO2-saving # / Area Reciever Reciever [%] Fluid Storage [ton/year] [m] Peak [hr]/[MWh] [kg/h]

624 / 120 m2 Cavity 55 MWT 92 Saturated Pressurized 6700 reciever 257 ᵒ C (Pressurized ) steel 40 bars steam accumulator 100,5 650 kW/m2 Water/Steam 1 (50% rate) / 15 100 000

Table 10. PS10-plant, Power Tower, Power cycle, Seville, Spain.

Power ηsteam Turbine Steamcycle ηSolar to Power Additional Thermalfluid

cycle cycle electricicty Generated Thermal Parabolic [%] [MWnom] [%] [GWh/year] Storage trough Rankine 30,7 Steam 250 degC 17 24,2 Molten Salt Synthetic oil 40 bar 11,02 2 Pressures 6,3 kV 8 MWh 50 Hz

87

Table 11. Solar-Tres, SENER power Tower data, Spain.

Heliostat Tower Design Data ηT Thermal Fluid Thermal CO2- savi # / Area Reciever Reciever Out/in Storage ng

[m] Peak [MWe] [%] [ºC] [hr]/[MWh] [ton/year] 2480 Cylindrical 120 >88 Molten salt 16 57 000 High flux 94 m2 120 565/290 647

Table 12. Solar-Tres, SENER power Tower data, Power cycle, Spain.

Power cycle ηsteam cycle Turbine ηSolar to electricicty Power Generated [%] [MW] [%] [GW/year] Rankine 38 17 >13,5 96,4

Table 13. Saguaro Organic Rankine Cycle, parabolic trough solar plant.

Parabolic Trough Design Data ηT Thermal Fluid # loops troughs/loop Plant [%] [ᵒC] 2 [m ] [MWe] [bar]

[kWT] 1 NA Mineral oil 6 843 315 10312

Table 13 continue: Saguaro Organic Rankine Cycle, parabolic trough solar plant.

Power cycle ηsteam cycle Turbine ηSolar to electricicty Power CO2-savings Generate [%] [%] d [ton/year] [GWh/year] Organic NA n-pentane 14,6 2 1 152 Rankine

88

3. TES –media

Table 14. Thermal characteristics of TES liquid media. (Pilkington Solar Int., 2000)

Storage material Temperature Average Average Average specific Volume specific Heat Liquid Media Cold Hot density conductivity heat capacity heat capacity ˚C ˚C [Kg/m3] [W/mK] [Kj/kgK] [KWh/m3] Mineral Oil 200 300 770 0,12 2,5 55 Synthetic Oil 250 350 900 0,11 2,3 57 Silicone Oil 300 400 900 0,1 2,1 52 Water 300 400 1000 NA 4,18 116 Salt Nitrite Salt 250 450 1825 0,57 1,5 152 Nitrate Salt 265 656 1870 0,52 1,6 250 Carbonate Salt 450 850 2100 2 1,8 430 Metal Na 270 530 850 71 1,3 80 NaK 270 530 855 25 0,9 NA

Table 15. Thermal Characteristics of liquid salt mixtures.

Storage Enthalpy material Transition change Average Liquid Media Temperature during melting density Singel Salts [˚C] [KJ/Kg] [Kg/m3]

Li2CO3 723 607 2110

Na2CO3 858 265 2500

K2CO3 898 200 2300

89

Table 16 Thermal characteristics, TES – solid media. (Pilkington Solar Int., 2000)

Storage material Temp erature Average Average Average specific Volume specific Solid Media Cold Hot density Heat conductivity Heat capacity Heat capacity ˚C ˚C [Kg/m3] [W/mK] [Kj/kgK] [KWh/m3] Sand rock mineral oil 200 300 1700 1 1,3 60 Reinforced concrete 200 400 2200 1,5 0,5 103,8 High temp. Concrete(DLR) 200 400 2750 1 0,916 140 Castable ceramic (DLR) 200 400 3500 1,4 0,866 168 NaCl (soild) 200 500 2160 7 0,85 150 Cast iron 200 400 7200 37 0,56 160 Cast steel 200 700 7800 40 0,6 450 Silica fire bricks 200 700 1820 1,5 1 150 Magnesia bricks 200 1200 3000 5 1,15 600

Table 17. LHTS – Thermal characteristics of PCM salt mixtures. (Tamme R., 2003)

PCM % of mix. Tmelting ∆Hu [ºC] [kJ/kg] NaOH / Na2CO3 x / 7,2 283 340

NaCl / NaOH) 26,8 / x 370 370 NaCl / KCl / LiCl x / 32,4 / 32,8 346 281 NaCl / NaNO3 /Na2SO4 5,7 / 85,5 / x 287 176 NaCl / NaNO3 x / 5,0 284 171

NaCl / NaNO3 5,0 / x 282 212 NaCl / KCl / MgCl2 42,5 / 20,5 / x 385-393 410 KNO3 / NaNO3 10 / x 290 170 KNO3 / KCl x / 4,5 320 150 KNO3 / KBr / KCl x / 4,7 / 7,3 3342 140

NaNO3 310 174

NaNo2 282 212

KNO3 337 116

90

91